ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1324 - (show annotations)
Fri May 10 11:40:06 2013 UTC (6 years, 6 months ago) by ph10
File size: 246870 byte(s)
Fix pcregrep so that it can find empty lines.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.33 28-April-2013
5 --------------------------
7 1. Added 'U' to some constants that are compared to unsigned integers, to
8 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
9 variables that are added to signed variables, to ensure the result is
10 signed and can be negated.
12 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
13 CMake config files.
15 3. Revise the creation of config.h.generic so that all boolean macros are
16 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
17 overriding via -D on the command line possible.
19 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
20 to unsigned int is reported to make a quite noticeable speed difference in
21 a specific Windows environment. Testing on Linux did also appear to show
22 some benefit (and it is clearly not harmful). Also fixed the definition of
23 Xop which should be unsigned.
25 5. Related to (4), changing the definition of the intermediate variable cc
26 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
27 improvements.
29 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
30 unnecessary spaces.
32 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
33 and later.
35 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
37 9. Optimizing fast_forward_start_bits in JIT.
39 10. Adding experimental support for callouts in JIT, and fixing some
40 issues revealed during this work. Namely:
42 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
44 (b) Minimum length was not checked before the matching is started.
46 11. The value of capture_last that is passed to callouts was incorrect in some
47 cases when there was a capture on one path that was subsequently abandoned
48 after a backtrack. Also, the capture_last value is now reset after a
49 recursion, since all captures are also reset in this case.
51 12. The interpreter no longer returns the "too many substrings" error in the
52 case when an overflowing capture is in a branch that is subsequently
53 abandoned after a backtrack.
55 13. In the pathological case when an offset vector of size 2 is used, pcretest
56 now prints out the matched string after a yield of 0 or 1.
58 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
59 Only supported by the JIT compiler at the moment.
61 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
63 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
64 offset of the starting point of the matching process, provided the offsets
65 vector is large enough.
67 17. The \A escape now records a lookbehind value of 1, though its execution
68 does not actually inspect the previous character. This is to ensure that,
69 in partial multi-segment matching, at least one character from the old
70 segment is retained when a new segment is processed. Otherwise, if there
71 are no lookbehinds in the pattern, \A might match incorrectly at the start
72 of a new segment.
74 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
76 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
77 gives some modest performance improvement in 8-bit mode.
79 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
80 be expressed in certain programming languages using Universal Character
81 Names.
83 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
84 which points out that "non characters" are not "characters that may not
85 appear in Unicode strings" but rather "characters that are reserved for
86 internal use and have only local meaning".
88 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
89 there was a conditional group that depended on an assertion, if the
90 assertion was false, the callout that immediately followed the alternation
91 in the condition was skipped when pcre_exec() was used for matching.
93 23. Allow an explicit callout to be inserted before an assertion that is the
94 condition for a conditional group, for compatibility with automatic
95 callouts, which always insert a callout at this point.
97 24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
98 confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
100 25. (*PRUNE) is now supported by the JIT compiler.
102 26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
104 27. Fix the case where there are two or more SKIPs with arguments that may be
105 ignored.
107 28. (*SKIP) is now supported by the JIT compiler.
109 29. (*THEN) is now supported by the JIT compiler.
111 30. Update RunTest with additional test selector options.
113 31. The way PCRE handles backtracking verbs has been changed in two ways.
115 (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
116 SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
117 backtracking. In some cases this makes it more Perl-compatible, but Perl's
118 rather obscure rules do not always do the same thing.
120 (2) Previously, backtracking verbs were confined within assertions. This is
121 no longer the case for positive assertions, except for (*ACCEPT). Again,
122 this sometimes improves Perl compatibility, and sometimes does not.
124 32. A number of tests that were in test 2 because Perl did things differently
125 have been moved to test 1, because either Perl or PCRE has changed, and
126 these tests are now compatible.
128 32. Control verbs are handled in the same way in JIT and interpreter.
130 33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
131 contained a forward subroutine reference caused a compile error.
133 34. Auto-detect and optimize limited repetitions in JIT.
135 35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
136 blocking (*UTF) etc.
138 36. In the interpreter, maximizing pattern repetitions for characters and
139 character types now use tail recursion, which reduces stack usage.
141 37. The value of the max lookbehind was not correctly preserved if a compiled
142 and saved regex was reloaded on a host of different endianness.
144 38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
145 of the compiled pattern block, expand the flags field from 16 to 32 bits
146 because it was almost full.
148 39. Try madvise first before posix_madvise.
150 40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
151 with a pattern such as ^$. It has taken 4 years for anybody to notice! The
152 original change locked out all matches of empty strings. This has been
153 changed so that one match of an empty string per line is recognized.
154 Subsequent searches on the same line (for colouring or for --only-matching,
155 for example) do not recognize empty strings.
158 Version 8.32 30-November-2012
159 -----------------------------
161 1. Improved JIT compiler optimizations for first character search and single
162 character iterators.
164 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
165 Patch by Daniel Richard G.
167 3. Single character iterator optimizations in the JIT compiler.
169 4. Improved JIT compiler optimizations for character ranges.
171 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
172 Reported by Giuseppe D'Angelo.
174 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
175 of a line, was being set incorrectly in cases where .* appeared inside
176 atomic brackets at the start of a pattern, or where there was a subsequent
177 *PRUNE or *SKIP.
179 7. Improved instruction cache flush for POWER/PowerPC.
180 Patch by Daniel Richard G.
182 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
183 grep:
185 (a) There is now no limit to the number of patterns to be matched.
187 (b) An error is given if a pattern is too long.
189 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
190 are now supported.
192 (d) --exclude-from and --include-from (multiple use) have been added.
194 (e) Exclusions and inclusions now apply to all files and directories, not
195 just to those obtained from scanning a directory recursively.
197 (f) Multiple uses of -f and --file-list are now supported.
199 (g) In a Windows environment, the default for -d has been changed from
200 "read" (the GNU grep default) to "skip", because otherwise the presence
201 of a directory in the file list provokes an error.
203 (h) The documentation has been revised and clarified in places.
205 9. Improve the matching speed of capturing brackets.
207 10. Changed the meaning of \X so that it now matches a Unicode extended
208 grapheme cluster.
210 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
211 out POSIX threads when JIT support is configured.
213 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
215 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
216 structure could go wrong in environments where size_t is not the same size
217 as int.
219 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
220 set.
222 15. The EBCDIC support had decayed; later updates to the code had included
223 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
224 general tidy up of EBCDIC-related issues, and the documentation was also
225 not quite right. There is now a test that can be run on ASCII systems to
226 check some of the EBCDIC-related things (but is it not a full test).
228 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
229 in a small tidy to the code.
231 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
233 18. If the --only-matching (-o) option in pcregrep is specified multiple
234 times, each one causes appropriate output. For example, -o1 -o2 outputs the
235 substrings matched by the 1st and 2nd capturing parentheses. A separating
236 string can be specified by --om-separator (default empty).
238 19. Improving the first n character searches.
240 20. Turn case lists for horizontal and vertical white space into macros so that
241 they are defined only once.
243 21. This set of changes together give more compatible Unicode case-folding
244 behaviour for characters that have more than one other case when UCP
245 support is available.
247 (a) The Unicode property table now has offsets into a new table of sets of
248 three or more characters that are case-equivalent. The MultiStage2.py
249 script that generates these tables (the pcre_ucd.c file) now scans
250 CaseFolding.txt instead of UnicodeData.txt for character case
251 information.
253 (b) The code for adding characters or ranges of characters to a character
254 class has been abstracted into a generalized function that also handles
255 case-independence. In UTF-mode with UCP support, this uses the new data
256 to handle characters with more than one other case.
258 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
259 whose other case is greater than 256 are now correctly matched
260 caselessly. Previously, the high codepoint matched the low one, but not
261 vice versa.
263 (d) The processing of \h, \H, \v, and \ in character classes now makes use
264 of the new class addition function, using character lists defined as
265 macros alongside the case definitions of 20 above.
267 (e) Caseless back references now work with characters that have more than
268 one other case.
270 (f) General caseless matching of characters with more than one other case
271 is supported.
273 22. Unicode character properties were updated from Unicode 6.2.0
275 23. Improved CMake support under Windows. Patch by Daniel Richard G.
277 24. Add support for 32-bit character strings, and UTF-32
279 25. Major JIT compiler update (code refactoring and bugfixing).
280 Experimental Sparc 32 support is added.
282 26. Applied a modified version of Daniel Richard G's patch to create
283 pcre.h.generic and config.h.generic by "make" instead of in the
284 PrepareRelease script.
286 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
287 pcre_compile.c when checking for a zero character.
289 28. Introducing a native interface for JIT. Through this interface, the compiled
290 machine code can be directly executed. The purpose of this interface is to
291 provide fast pattern matching, so several sanity checks are not performed.
292 However, feature tests are still performed. The new interface provides
293 1.4x speedup compared to the old one.
295 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
296 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
297 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
299 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
300 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
301 the "old" RFC 2279). Instead, it ended up passing a negative length to
302 pcre_exec().
304 31. Add support for GCC's visibility feature to hide internal functions.
306 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
307 "unknown -C option" after outputting 0 or 1.
309 33. There is now support for generating a code coverage report for the test
310 suite in environments where gcc is the compiler and lcov is installed. This
311 is mainly for the benefit of the developers.
313 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
314 unaddressable using valgrind annotations, allowing valgrind to detect
315 invalid memory accesses. This is mainly for the benefit of the developers.
317 25. (*UTF) can now be used to start a pattern in any of the three libraries.
319 26. Give configure error if --enable-cpp but no C++ compiler found.
322 Version 8.31 06-July-2012
323 -------------------------
325 1. Fixing a wrong JIT test case and some compiler warnings.
327 2. Removed a bashism from the RunTest script.
329 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
330 to unsigned type, result still unsigned" that was given by an MS compiler
331 on encountering the code "-sizeof(xxx)".
333 4. Partial matching support is added to the JIT compiler.
335 5. Fixed several bugs concerned with partial matching of items that consist
336 of more than one character:
338 (a) /^(..)\1/ did not partially match "aba" because checking references was
339 done on an "all or nothing" basis. This also applied to repeated
340 references.
342 (b) \R did not give a hard partial match if \r was found at the end of the
343 subject.
345 (c) \X did not give a hard partial match after matching one or more
346 characters at the end of the subject.
348 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
349 a partial match for the string "\r".
351 (e) When newline was set to CRLF, the metacharacter "." did not recognize
352 a partial match for a CR character at the end of the subject string.
354 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
355 running pcretest, the text "(JIT)" added to the output whenever JIT is
356 actually used to run the match.
358 7. Individual JIT compile options can be set in pcretest by following -s+[+]
359 or /S+[+] with a digit between 1 and 7.
361 8. OP_NOT now supports any UTF character not just single-byte ones.
363 9. (*MARK) control verb is now supported by the JIT compiler.
365 10. The command "./RunTest list" lists the available tests without actually
366 running any of them. (Because I keep forgetting what they all are.)
370 12. Applied a (slightly modified) user-supplied patch that improves performance
371 when the heap is used for recursion (compiled with --disable-stack-for-
372 recursion). Instead of malloc and free for each heap frame each time a
373 logical recursion happens, frames are retained on a chain and re-used where
374 possible. This sometimes gives as much as 30% improvement.
376 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
377 call.
379 14. As documented, (*COMMIT) is now confined to within a positive assertion.
381 15. It is now possible to link pcretest with libedit as an alternative to
382 libreadline.
384 16. (*COMMIT) control verb is now supported by the JIT compiler.
386 17. The Unicode data tables have been updated to Unicode 6.1.0.
388 18. Added --file-list option to pcregrep.
390 19. Added binary file support to pcregrep, including the -a, --binary-files,
391 -I, and --text options.
393 20. The madvise function is renamed for posix_madvise for QNX compatibility
394 reasons. Fixed by Giuseppe D'Angelo.
396 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
398 22. Changed the option for creating symbolic links for 16-bit man pages from
399 -s to -sf so that re-installing does not cause issues.
401 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
403 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
404 matches in certain environments (the workspace was not being correctly
405 retained). Also added to pcre_dfa_exec() a simple plausibility check on
406 some of the workspace data at the beginning of a restart.
408 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
409 was not doing so when it should - probably a typo introduced by SVN 528
410 (change 8.10/14).
412 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
413 \w+ when the character tables indicated that \x{c4} was a word character.
414 There were several related cases, all because the tests for doing a table
415 lookup were testing for characters less than 127 instead of 255.
417 27. If a pattern contains capturing parentheses that are not used in a match,
418 their slots in the ovector are set to -1. For those that are higher than
419 any matched groups, this happens at the end of processing. In the case when
420 there were back references that the ovector was too small to contain
421 (causing temporary malloc'd memory to be used during matching), and the
422 highest capturing number was not used, memory off the end of the ovector
423 was incorrectly being set to -1. (It was using the size of the temporary
424 memory instead of the true size.)
426 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
427 ovector size, it uses memory at the end of the block that it has got.
429 29. Check for an overlong MARK name and give an error at compile time. The
430 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
432 30. JIT compiler update.
434 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
435 Rill for the patch.
437 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
439 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
441 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
442 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
443 was enabled.
445 35. Improve JIT code generation for greedy plus quantifier.
447 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
448 1 to "aa" instead of to an empty string. The bug affected repeated groups
449 that could potentially match an empty string.
451 37. Optimizing single character iterators in JIT.
453 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
454 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
455 codepoints that are too big for the mode are faulted, and in a UTF mode,
456 disallowed codepoints are also faulted.
458 39. If PCRE was compiled with UTF support, in three places in the DFA
459 matcher there was code that should only have been obeyed in UTF mode, but
460 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
461 processing when bytes with values greater than 127 were present. In 16-bit
462 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
463 both cases the values are those that cannot be the first data item in a UTF
464 character. The three items that might have provoked this were recursions,
465 possessively repeated groups, and atomic groups.
467 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
468 and pcregrep, because some OS require shared objects to be explicitly
469 passed to ld, causing the link step to fail if they are not.
471 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
472 mode, patterns that started with \h* or \R* might be incorrectly matched.
475 Version 8.30 04-February-2012
476 -----------------------------
478 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
479 name is defined in ctype.h.
481 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
482 only in quite long subpatterns.
484 3. Removed the function pcre_info(), which has been obsolete and deprecated
485 since it was replaced by pcre_fullinfo() in February 2000.
487 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
488 match a (*MARK), and the match failed at the start of the subject, a
489 reference to memory before the start of the subject could occur. This bug
490 was introduced by fix 17 of release 8.21.
492 5. A reference to an unset group with zero minimum repetition was giving
493 totally wrong answers (in non-JavaScript-compatibility mode). For example,
494 /(another)?(\1?)test/ matched against "hello world test". This bug was
495 introduced in release 8.13.
497 6. Add support for 16-bit character strings (a large amount of work involving
498 many changes and refactorings).
500 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
501 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
502 from a file.
504 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
505 rounding is not applied in this particular case).
507 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
508 if they appear, or are escaped, in patterns.
510 10. Get rid of a number of -Wunused-but-set-variable warnings.
512 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
513 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
514 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
515 also returns the mark "x". This bug applied to capturing parentheses,
516 non-capturing parentheses, and atomic parentheses. It also applied to some
517 assertions.
519 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
520 information out of configure.ac instead of relying on pcre.h.generic, which
521 is not stored in the repository.
523 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
524 -lreadline.
526 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
528 15. Applied Graycode's patch to put the top-level frame on the stack rather
529 than the heap when not using the stack for recursion. This gives a
530 performance improvement in many cases when recursion is not deep.
532 16. Experimental code added to "pcretest -C" to output the stack frame size.
535 Version 8.21 12-Dec-2011
536 ------------------------
538 1. Updating the JIT compiler.
540 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
541 are added as well.
543 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
544 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
545 calling _pcre_jit_exec. Some extra comments are added.
547 4. (*MARK) settings inside atomic groups that do not contain any capturing
548 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
549 was introduced by change 18 for 8.20.
551 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
552 ECMA-262 standard.
554 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
555 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
556 This bug was probably introduced by change 9 of 8.13.
558 7. While fixing 6 above, I noticed that a number of other items were being
559 incorrectly rejected as "not fixed length". This arose partly because newer
560 opcodes had not been added to the fixed-length checking code. I have (a)
561 corrected the bug and added tests for these items, and (b) arranged for an
562 error to occur if an unknown opcode is encountered while checking for fixed
563 length instead of just assuming "not fixed length". The items that were
564 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
565 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
566 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
568 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
569 being incorrectly compiled and would have given unpredicatble results.
571 9. A possessively repeated subpattern with minimum repeat count greater than
572 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
573 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
574 the first (A) could occur when it should not.
576 10. Add a cast and remove a redundant test from the code.
578 11. JIT should use pcre_malloc/pcre_free for allocation.
580 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
581 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
582 is anything other than /usr, -L is still shown).
584 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
586 14. Perl does not support \N without a following name in a [] class; PCRE now
587 also gives an error.
589 15. If a forward reference was repeated with an upper limit of around 2000,
590 it caused the error "internal error: overran compiling workspace". The
591 maximum number of forward references (including repeats) was limited by the
592 internal workspace, and dependent on the LINK_SIZE. The code has been
593 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
594 the default depends on LINK_SIZE. There is a new upper limit (for safety)
595 of around 200,000 forward references. While doing this, I also speeded up
596 the filling in of repeated forward references.
598 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
599 incorrectly expecting the subject to contain another "a" after the start.
601 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
602 in the match, the SKIP should be ignored. This was not happening; instead
603 the SKIP was being treated as NOMATCH. For patterns such as
604 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
605 tested.
607 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
608 now much more compatible with Perl, in particular in cases where the result
609 is a non-match for a non-anchored pattern. For example, if
610 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
611 "m", where previously it did not return a name. A side effect of this
612 change is that for partial matches, the last encountered mark name is
613 returned, as for non matches. A number of tests that were previously not
614 Perl-compatible have been moved into the Perl-compatible test files. The
615 refactoring has had the pleasing side effect of removing one argument from
616 the match() function, thus reducing its stack requirements.
618 19. If the /S+ option was used in pcretest to study a pattern using JIT,
619 subsequent uses of /S (without +) incorrectly behaved like /S+.
621 21. Retrieve executable code size support for the JIT compiler and fixing
622 some warnings.
624 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
625 not work when the shorter character appeared right at the end of the
626 subject string.
628 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
629 systems.
631 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
632 output it when the /M option is used in pcretest.
634 25. The CheckMan script was not being included in the distribution. Also, added
635 an explicit "perl" to run Perl scripts from the PrepareRelease script
636 because this is reportedly needed in Windows.
638 26. If study data was being save in a file and studying had not found a set of
639 "starts with" bytes for the pattern, the data written to the file (though
640 never used) was taken from uninitialized memory and so caused valgrind to
641 complain.
643 27. Updated RunTest.bat as provided by Sheri Pierce.
645 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
647 29. Computation of memory usage for the table of capturing group names was
648 giving an unnecessarily large value.
651 Version 8.20 21-Oct-2011
652 ------------------------
654 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
655 a POSIX class. After further experiments with Perl, which convinced me that
656 Perl has bugs and confusions, a closing square bracket is no longer allowed
657 in a POSIX name. This bug also affected patterns with classes that started
658 with full stops.
660 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
661 captured substring, but while checking the failing first alternative,
662 substring 1 is temporarily captured. If the output vector supplied to
663 pcre_exec() was not big enough for this capture, the yield of the function
664 was still zero ("insufficient space for captured substrings"). This cannot
665 be totally fixed without adding another stack variable, which seems a lot
666 of expense for a edge case. However, I have improved the situation in cases
667 such as /(a)(b)x|abc/ matched against "abc", where the return code
668 indicates that fewer than the maximum number of slots in the ovector have
669 been set.
671 3. Related to (2) above: when there are more back references in a pattern than
672 slots in the output vector, pcre_exec() uses temporary memory during
673 matching, and copies in the captures as far as possible afterwards. It was
674 using the entire output vector, but this conflicts with the specification
675 that only 2/3 is used for passing back captured substrings. Now it uses
676 only the first 2/3, for compatibility. This is, of course, another edge
677 case.
679 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
680 main code base, and can be used by building with --enable-jit. When this is
681 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
682 runtime --no-jit option is given.
684 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
685 ovector, the return from the function was zero, implying that there were
686 other matches that did not fit. The correct "exactly full" value is now
687 returned.
689 6. If a subpattern that was called recursively or as a subroutine contained
690 (*PRUNE) or any other control that caused it to give a non-standard return,
691 invalid errors such as "Error -26 (nested recursion at the same subject
692 position)" or even infinite loops could occur.
694 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
695 computing the minimum length on reaching *ACCEPT, and so ended up with the
696 wrong value of 1 rather than 0. Further investigation indicates that
697 computing a minimum subject length in the presence of *ACCEPT is difficult
698 (think back references, subroutine calls), and so I have changed the code
699 so that no minimum is registered for a pattern that contains *ACCEPT.
701 8. If (*THEN) was present in the first (true) branch of a conditional group,
702 it was not handled as intended. [But see 16 below.]
704 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
705 Sheri Pierce.
707 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
708 the first byte in a match must be "a".
710 11. Change 17 for 8.13 increased the recursion depth for patterns like
711 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
712 pattern contains any instances of (*THEN). If it does not, the old
713 optimizations are restored. It would be nice to do this on a per-group
714 basis, but at the moment that is not feasible.
716 12. In some environments, the output of pcretest -C is CRLF terminated. This
717 broke RunTest's code that checks for the link size. A single white space
718 character after the value is now allowed for.
720 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
721 For "fr", it uses the Windows-specific input and output files.
723 14. If (*THEN) appeared in a group that was called recursively or as a
724 subroutine, it did not work as intended. [But see next item.]
726 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
727 pattern fragments (but not containing any | characters). If A and B are
728 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
729 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
730 D. In other words, Perl considers parentheses that do not contain any |
731 characters to be part of a surrounding alternative, whereas PCRE was
732 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
733 differently. PCRE now behaves in the same way as Perl, except in the case
734 of subroutine/recursion calls such as (?1) which have in any case always
735 been different (but PCRE had them first :-).
737 16. Related to 15 above: Perl does not treat the | in a conditional group as
738 creating alternatives. Such a group is treated in the same way as an
739 ordinary group without any | characters when processing (*THEN). PCRE has
740 been changed to match Perl's behaviour.
742 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
743 RunGrepTest script failed.
745 18. Change 22 for version 13 caused atomic groups to use more stack. This is
746 inevitable for groups that contain captures, but it can lead to a lot of
747 stack use in large patterns. The old behaviour has been restored for atomic
748 groups that do not contain any capturing parentheses.
750 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
751 suppress the check for a minimum subject length at run time. (If it was
752 given to pcre_exec() or pcre_dfa_exec() it did work.)
754 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
755 fail to work when decoding hex characters in data strings in EBCDIC
756 environments.
758 21. It appears that in at least one Mac OS environment, the isxdigit() function
759 is implemented as a macro that evaluates to its argument more than once,
760 contravening the C 90 Standard (I haven't checked a later standard). There
761 was an instance in pcretest which caused it to go wrong when processing
762 \x{...} escapes in subject strings. The has been rewritten to avoid using
763 things like p++ in the argument of isxdigit().
766 Version 8.13 16-Aug-2011
767 ------------------------
769 1. The Unicode data tables have been updated to Unicode 6.0.0.
771 2. Two minor typos in pcre_internal.h have been fixed.
773 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
774 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
775 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
777 4. There were a number of related bugs in the code for matching backrefences
778 caselessly in UTF-8 mode when codes for the characters concerned were
779 different numbers of bytes. For example, U+023A and U+2C65 are an upper
780 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
781 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
782 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
783 2-byte code at the end of the subject (it thought there wasn't enough data
784 left).
786 5. Comprehensive information about what went wrong is now returned by
787 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
788 as the output vector has at least 2 elements. The offset of the start of
789 the failing character and a reason code are placed in the vector.
791 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
792 now returned is for the first byte of the failing character, instead of the
793 last byte inspected. This is an incompatible change, but I hope it is small
794 enough not to be a problem. It makes the returned offset consistent with
795 pcre_exec() and pcre_dfa_exec().
797 7. pcretest now gives a text phrase as well as the error number when
798 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
799 failure, the offset and reason code are output.
801 8. When \R was used with a maximizing quantifier it failed to skip backwards
802 over a \r\n pair if the subsequent match failed. Instead, it just skipped
803 back over a single character (\n). This seems wrong (because it treated the
804 two characters as a single entity when going forwards), conflicts with the
805 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
806 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
807 has been changed.
809 9. Some internal refactoring has changed the processing so that the handling
810 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
811 time (the PCRE_DOTALL option was changed this way some time ago: version
812 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
813 which was always a bit of a fudge. It also means that there is one less
814 argument for the match() function, which reduces its stack requirements
815 slightly. This change also fixes an incompatibility with Perl: the pattern
816 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
818 10. More internal refactoring has drastically reduced the number of recursive
819 calls to match() for possessively repeated groups such as (abc)++ when
820 using pcre_exec().
822 11. While implementing 10, a number of bugs in the handling of groups were
823 discovered and fixed:
825 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
826 (a|)*(?1) gave a compile-time internal error.
827 ((a|)+)+ did not notice that the outer group could match an empty string.
828 (^a|^)+ was not marked as anchored.
829 (.*a|.*)+ was not marked as matching at start or after a newline.
831 12. Yet more internal refactoring has removed another argument from the match()
832 function. Special calls to this function are now indicated by setting a
833 value in a variable in the "match data" data block.
835 13. Be more explicit in pcre_study() instead of relying on "default" for
836 opcodes that mean there is no starting character; this means that when new
837 ones are added and accidentally left out of pcre_study(), testing should
838 pick them up.
840 14. The -s option of pcretest has been documented for ages as being an old
841 synonym of -m (show memory usage). I have changed it to mean "force study
842 for every regex", that is, assume /S for every regex. This is similar to -i
843 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
844 using it. It makes it easier to run collections of tests with and without
845 study enabled, and thereby test pcre_study() more easily. All the standard
846 tests are now run with and without -s (but some patterns can be marked as
847 "never study" - see 20 below).
849 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
850 restoration of the capturing data to the outer values was not happening
851 correctly.
853 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
854 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
855 pattern had matched an empty string, and so incorrectly returned a no
856 match.
858 17. There was optimizing code for the last branch of non-capturing parentheses,
859 and also for the obeyed branch of a conditional subexpression, which used
860 tail recursion to cut down on stack usage. Unfortunately, now that there is
861 the possibility of (*THEN) occurring in these branches, tail recursion is
862 no longer possible because the return has to be checked for (*THEN). These
863 two optimizations have therefore been removed. [But see 8.20/11 above.]
865 18. If a pattern containing \R was studied, it was assumed that \R always
866 matched two bytes, thus causing the minimum subject length to be
867 incorrectly computed because \R can also match just one byte.
869 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
870 was incorrectly computed.
872 20. If /S is present twice on a test pattern in pcretest input, it now
873 *disables* studying, thereby overriding the use of -s on the command line
874 (see 14 above). This is necessary for one or two tests to keep the output
875 identical in both cases.
877 21. When (*ACCEPT) was used in an assertion that matched an empty string and
878 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
880 22. When an atomic group that contained a capturing parenthesis was
881 successfully matched, but the branch in which it appeared failed, the
882 capturing was not being forgotten if a higher numbered group was later
883 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
884 group 1 to "a", when in fact it should be unset. This applied to multi-
885 branched capturing and non-capturing groups, repeated or not, and also to
886 positive assertions (capturing in negative assertions does not happen
887 in PCRE) and also to nested atomic groups.
889 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
890 subject after a captured substring, to make it easier to tell which of a
891 number of identical substrings has been captured.
893 24. The way atomic groups are processed by pcre_exec() has been changed so that
894 if they are repeated, backtracking one repetition now resets captured
895 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
896 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
897 "aaa". Previously, it would have been "a". As part of this code
898 refactoring, the way recursive calls are handled has also been changed.
900 25. If an assertion condition captured any substrings, they were not passed
901 back unless some other capturing happened later. For example, if
902 (?(?=(a))a) was matched against "a", no capturing was returned.
904 26. When studying a pattern that contained subroutine calls or assertions,
905 the code for finding the minimum length of a possible match was handling
906 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
907 group 1 called group 2 while simultaneously a separate group 2 called group
908 1). A stack overflow occurred in this case. I have fixed this by limiting
909 the recursion depth to 10.
911 27. Updated RunTest.bat in the distribution to the version supplied by Tom
912 Fortmann. This supports explicit test numbers on the command line, and has
913 argument validation and error reporting.
915 28. An instance of \X with an unlimited repeat could fail if at any point the
916 first character it looked at was a mark character.
918 29. Some minor code refactoring concerning Unicode properties and scripts
919 should reduce the stack requirement of match() slightly.
921 30. Added the '=' option to pcretest to check the setting of unused capturing
922 slots at the end of the pattern, which are documented as being -1, but are
923 not included in the return count.
925 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
926 compiled something random. Now it gives a compile-time error (as does
927 Perl).
929 32. A *MARK encountered during the processing of a positive assertion is now
930 recorded and passed back (compatible with Perl).
932 33. If --only-matching or --colour was set on a pcregrep call whose pattern
933 had alternative anchored branches, the search for a second match in a line
934 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
935 matched the line "0102" twice. The same bug affected patterns that started
936 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
937 twice.
939 34. Previously, PCRE did not allow quantification of assertions. However, Perl
940 does, and because of capturing effects, quantifying parenthesized
941 assertions may at times be useful. Quantifiers are now allowed for
942 parenthesized assertions.
944 35. A minor code tidy in pcre_compile() when checking options for \R usage.
946 36. \g was being checked for fancy things in a character class, when it should
947 just be a literal "g".
949 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
950 appearance of a nested POSIX class supersedes an apparent external class.
951 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
952 unescaped square brackets may also appear as part of class names. For
953 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
954 more like Perl. (But see 8.20/1 above.)
956 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
957 was because it thought it was \N{name}, which is not supported).
959 39. Add minix to OS list not supporting the -S option in pcretest.
961 40. PCRE tries to detect cases of infinite recursion at compile time, but it
962 cannot analyze patterns in sufficient detail to catch mutual recursions
963 such as ((?1))((?2)). There is now a runtime test that gives an error if a
964 subgroup is called recursively as a subpattern for a second time at the
965 same position in the subject string. In previous releases this might have
966 been caught by the recursion limit, or it might have run out of stack.
968 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
969 happen only once. PCRE was, however incorrectly giving a compile time error
970 "recursive call could loop indefinitely" because it cannot analyze the
971 pattern in sufficient detail. The compile time test no longer happens when
972 PCRE is compiling a conditional subpattern, but actual runaway loops are
973 now caught at runtime (see 40 above).
975 42. It seems that Perl allows any characters other than a closing parenthesis
976 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
977 has been changed to be the same.
979 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
980 as not to get warnings when autogen.sh is called. Also changed
981 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
983 44. To help people who use pcregrep to scan files containing exceedingly long
984 lines, the following changes have been made:
986 (a) The default value of the buffer size parameter has been increased from
987 8K to 20K. (The actual buffer used is three times this size.)
989 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
990 PCRE is built.
992 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
993 to be set at run time.
995 (d) Numerical values in pcregrep options can be followed by K or M, for
996 example --buffer-size=50K.
998 (e) If a line being scanned overflows pcregrep's buffer, an error is now
999 given and the return code is set to 2.
1001 45. Add a pointer to the latest mark to the callout data block.
1003 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1004 partial match of an empty string instead of no match. This was specific to
1005 the use of ".".
1007 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1008 complete match instead of a partial match. This bug was dependent on both
1009 the PCRE_UTF8 and PCRE_DOTALL options being set.
1011 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1012 starting byte set, because \b was not being ignored.
1015 Version 8.12 15-Jan-2011
1016 ------------------------
1018 1. Fixed some typos in the markup of the man pages, and wrote a script that
1019 checks for such things as part of the documentation building process.
1021 2. On a big-endian 64-bit system, pcregrep did not correctly process the
1022 --match-limit and --recursion-limit options (added for 8.11). In
1023 particular, this made one of the standard tests fail. (The integer value
1024 went into the wrong half of a long int.)
1026 3. If the --colour option was given to pcregrep with -v (invert match), it
1027 did strange things, either producing crazy output, or crashing. It should,
1028 of course, ignore a request for colour when reporting lines that do not
1029 match.
1031 4. Another pcregrep bug caused similar problems if --colour was specified with
1032 -M (multiline) and the pattern match finished with a line ending.
1034 5. In pcregrep, when a pattern that ended with a literal newline sequence was
1035 matched in multiline mode, the following line was shown as part of the
1036 match. This seems wrong, so I have changed it.
1038 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
1039 the check for further matches in the same line (so they could be coloured)
1040 to overrun the end of the current line. If another match was found, it was
1041 incorrectly shown (and then shown again when found in the next line).
1043 7. If pcregrep was compiled under Windows, there was a reference to the
1044 function pcregrep_exit() before it was defined. I am assuming this was
1045 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1046 reported by a user. I've moved the definition above the reference.
1049 Version 8.11 10-Dec-2010
1050 ------------------------
1052 1. (*THEN) was not working properly if there were untried alternatives prior
1053 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1054 backtracked to try for "b" instead of moving to the next alternative branch
1055 at the same level (in this case, to look for "c"). The Perl documentation
1056 is clear that when (*THEN) is backtracked onto, it goes to the "next
1057 alternative in the innermost enclosing group".
1059 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1060 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1061 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1062 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1063 (*THEN).
1065 3. If \s appeared in a character class, it removed the VT character from
1066 the class, even if it had been included by some previous item, for example
1067 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1068 of \s, but is part of the POSIX "space" class.)
1070 4. A partial match never returns an empty string (because you can always
1071 match an empty string at the end of the subject); however the checking for
1072 an empty string was starting at the "start of match" point. This has been
1073 changed to the "earliest inspected character" point, because the returned
1074 data for a partial match starts at this character. This means that, for
1075 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1076 (previously it gave "no match").
1078 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1079 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1080 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1081 has an implication that the given string is incomplete (because a partial
1082 match is preferred over a full match). For this reason, these items now
1083 give a partial match in this situation. [Aside: previously, the one case
1084 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1085 match rather than a full match, which was wrong by the old rules, but is
1086 now correct.]
1088 6. There was a bug in the handling of #-introduced comments, recognized when
1089 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1090 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1091 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1092 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1093 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1094 places in pcre_compile().
1096 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1097 comments when looking ahead for named forward references to subpatterns,
1098 the only newline sequence it recognized was NL. It now handles newlines
1099 according to the set newline convention.
1101 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1102 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1103 cater for a lack of strerror(). These oversights have been fixed.
1105 9. Added --match-limit and --recursion-limit to pcregrep.
1107 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1109 11. When the -o option was used, pcregrep was setting a return code of 1, even
1110 when matches were found, and --line-buffered was not being honoured.
1112 12. Added an optional parentheses number to the -o and --only-matching options
1113 of pcregrep.
1115 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1116 can match an empty string. The code to do it in pcretest and pcredemo
1117 needed fixing:
1119 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1120 only one byte after an empty string match just before CRLF (this case
1121 just got forgotten; "any" and "anycrlf" were OK).
1123 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1124 mode when an empty string match preceded an ASCII character followed by
1125 a non-ASCII character. (The code for advancing by one character rather
1126 than one byte was nonsense.)
1128 (c) The pcredemo.c sample program did not have any code at all to handle
1129 the cases when CRLF is a valid newline sequence.
1131 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1132 as a starting offset was within the subject string. There is now a new
1133 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1134 negative or greater than the length of the string. In order to test this,
1135 pcretest is extended to allow the setting of negative starting offsets.
1137 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1138 starting offset points to the beginning of a UTF-8 character was
1139 unnecessarily clumsy. I tidied it up.
1141 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1142 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1144 17. Nobody had reported that the --include_dir option, which was added in
1145 release 7.7 should have been called --include-dir (hyphen, not underscore)
1146 for compatibility with GNU grep. I have changed it to --include-dir, but
1147 left --include_dir as an undocumented synonym, and the same for
1148 --exclude-dir, though that is not available in GNU grep, at least as of
1149 release 2.5.4.
1151 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1152 characters from a string of bytes) have been redefined so as not to use
1153 loops, in order to improve performance in some environments. At the same
1154 time, I abstracted some of the common code into auxiliary macros to save
1155 repetition (this should not affect the compiled code).
1157 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1158 compile-time error is now given if \c is not followed by an ASCII
1159 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1160 different, and any byte value is allowed.)
1162 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1163 START_OPTIMIZE option, which is now allowed at compile time - but just
1164 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1165 to pcregrep and other applications that have no direct access to PCRE
1166 options. The new /Y option in pcretest sets this option when calling
1167 pcre_compile().
1169 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1170 back references. Groups containing recursive back references were forced to
1171 be atomic by that change, but in the case of named groups, the amount of
1172 memory required was incorrectly computed, leading to "Failed: internal
1173 error: code overflow". This has been fixed.
1175 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1176 pcretest.c, to avoid build problems in some Borland environments.
1179 Version 8.10 25-Jun-2010
1180 ------------------------
1182 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1183 THEN.
1185 2. (*ACCEPT) was not working when inside an atomic group.
1187 3. Inside a character class, \B is treated as a literal by default, but
1188 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1189 causes the error). The code is unchanged, but I tidied the documentation.
1191 4. Inside a character class, PCRE always treated \R and \X as literals,
1192 whereas Perl faults them if its -w option is set. I have changed PCRE so
1193 that it faults them when PCRE_EXTRA is set.
1195 5. Added support for \N, which always matches any character other than
1196 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1198 6. When compiling pcregrep with newer versions of gcc which may have
1199 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1200 declared with attribute warn_unused_result" were given. Just casting the
1201 result to (void) does not stop the warnings; a more elaborate fudge is
1202 needed. I've used a macro to implement this.
1204 7. Minor change to pcretest.c to avoid a compiler warning.
1206 8. Added four artifical Unicode properties to help with an option to make
1207 \s etc use properties (see next item). The new properties are: Xan
1208 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1210 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1211 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1212 this option. Modified pcretest to add /W to test this facility. Added
1213 REG_UCP to make it available via the POSIX interface.
1215 10. Added --line-buffered to pcregrep.
1217 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1218 studied, and the match started with a letter with a code point greater than
1219 127 whose first byte was different to the first byte of the other case of
1220 the letter, the other case of this starting letter was not recognized
1221 (#976).
1223 12. If a pattern that was studied started with a repeated Unicode property
1224 test, for example, \p{Nd}+, there was the theoretical possibility of
1225 setting up an incorrect bitmap of starting bytes, but fortunately it could
1226 not have actually happened in practice until change 8 above was made (it
1227 added property types that matched character-matching opcodes).
1229 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1230 possible starting bytes for non-anchored patterns.
1232 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1233 \R, and also a number of cases that involve Unicode properties, both
1234 explicit and implicit when PCRE_UCP is set.
1236 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1237 input, it could crash or give wrong results if characters with values
1238 greater than 0xc0 were present in the subject string. (Detail: it assumed
1239 UTF-8 input when processing these items.)
1241 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1242 size_t is 64-bit (#991).
1244 17. Added a check for running out of memory when PCRE is compiled with
1245 --disable-stack-for-recursion (#990).
1247 18. If the last data line in a file for pcretest does not have a newline on
1248 the end, a newline was missing in the output.
1250 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1251 less than 128) in its various bitmaps. However, there is a facility for
1252 generating tables according to the current locale when PCRE is compiled. It
1253 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1254 characters, are recognized by isspace() and therefore were getting set in
1255 these tables, and indeed these tables seem to approximate to ISO 8859. This
1256 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1257 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1258 which of course cannot start UTF-8 characters. I have changed the code so
1259 that only real ASCII characters (less than 128) and the correct starting
1260 bytes for UTF-8 encodings are set for characters greater than 127 when in
1261 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1262 altogether.)
1264 20. Added the /T option to pcretest so as to be able to run tests with non-
1265 standard character tables, thus making it possible to include the tests
1266 used for 19 above in the standard set of tests.
1268 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1269 reference to a subpattern the other side of a comment that contains an
1270 opening parenthesis caused either an internal compiling error, or a
1271 reference to the wrong subpattern.
1274 Version 8.02 19-Mar-2010
1275 ------------------------
1277 1. The Unicode data tables have been updated to Unicode 5.2.0.
1279 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1280 configured.
1282 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1283 original author of that file, following a query about its status.
1285 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1286 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1288 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1289 quantifier applied to a forward-referencing subroutine call, could compile
1290 incorrect code or give the error "internal error: previously-checked
1291 referenced subpattern not found".
1293 6. Both MS Visual Studio and Symbian OS have problems with initializing
1294 variables to point to external functions. For these systems, therefore,
1295 pcre_malloc etc. are now initialized to local functions that call the
1296 relevant global functions.
1298 7. There were two entries missing in the vectors called coptable and poptable
1299 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1300 I've fixed the data, and added a kludgy way of testing at compile time that
1301 the lengths are correct (equal to the number of opcodes).
1303 8. Following on from 7, I added a similar kludge to check the length of the
1304 eint vector in pcreposix.c.
1306 9. Error texts for pcre_compile() are held as one long string to avoid too
1307 much relocation at load time. To find a text, the string is searched,
1308 counting zeros. There was no check for running off the end of the string,
1309 which could happen if a new error number was added without updating the
1310 string.
1312 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1314 11. \K was not working if it appeared in an atomic group or in a group that
1315 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1316 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1317 the assertion is positive, but not if it is negative.
1319 12. Change 11 fortuitously reduced the size of the stack frame used in the
1320 "match()" function of pcre_exec.c by one pointer. Forthcoming
1321 implementation of support for (*MARK) will need an extra pointer on the
1322 stack; I have reserved it now, so that the stack frame size does not
1323 decrease.
1325 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1326 item in branch that calls a recursion is a subroutine call - as in the
1327 second branch in the above example - was incorrectly given the compile-
1328 time error "recursive call could loop indefinitely" because pcre_compile()
1329 was not correctly checking the subroutine for matching a non-empty string.
1331 14. The checks for overrunning compiling workspace could trigger after an
1332 overrun had occurred. This is a "should never occur" error, but it can be
1333 triggered by pathological patterns such as hundreds of nested parentheses.
1334 The checks now trigger 100 bytes before the end of the workspace.
1336 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1339 Version 8.01 19-Jan-2010
1340 ------------------------
1342 1. If a pattern contained a conditional subpattern with only one branch (in
1343 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1344 computed the wrong minimum data length (which is of course zero for such
1345 subpatterns). This could cause incorrect "no match" results.
1347 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1348 the pattern is reset in the first branch, pcre_compile() failed with
1349 "internal error: code overflow at offset...". This happened only when
1350 the reset was to the original external option setting. (An optimization
1351 abstracts leading options settings into an external setting, which was the
1352 cause of this.)
1354 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1355 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1356 assertion pattern did not match (meaning that the assertion was true), it
1357 was incorrectly treated as false if the SKIP had been reached during the
1358 matching. This also applied to assertions used as conditions.
1360 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1361 assertion subpattern, including such a pattern used as a condition,
1362 unpredictable results occurred, instead of the error return
1365 5. The C++ GlobalReplace function was not working like Perl for the special
1366 situation when an empty string is matched. It now does the fancy magic
1367 stuff that is necessary.
1369 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1370 removed. (These were left over from very, very early versions of PCRE.)
1372 7. Some cosmetic changes to the code to make life easier when compiling it
1373 as part of something else:
1375 (a) Change DEBUG to PCRE_DEBUG.
1377 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1378 called "current" as "current_branch", to prevent a collision with the
1379 Linux macro when compiled as a kernel module.
1381 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1382 prevent a collision with the Linux macro when compiled as a kernel
1383 module.
1385 8. In pcre_compile() there are some checks for integer overflows that used to
1386 cast potentially large values to (double). This has been changed to that
1387 when building, a check for int64_t is made, and if it is found, it is used
1388 instead, thus avoiding the use of floating point arithmetic. (There is no
1389 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1390 double.
1392 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1393 2005 (difference between two addresses compared to an unsigned value).
1395 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1396 custom one, because of the following reported problem in Windows:
1398 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1399 under Win32.
1400 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1401 therefore missing the function definition.
1402 - The compiler thus generates a "C" signature for the test function.
1403 - The linker fails to find the "C" function.
1404 - PCRE fails to configure if asked to do so against libbz2.
1406 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1407 messages were output:
1409 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1410 rerunning libtoolize, to keep the correct libtool macros in-tree.
1411 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1413 I have done both of these things.
1415 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1416 most of the time, it *can* run out if it is given a pattern that contains a
1417 runaway infinite recursion. I updated the discussion in the pcrestack man
1418 page.
1420 13. Now that we have gone to the x.xx style of version numbers, the minor
1421 version may start with zero. Using 08 or 09 is a bad idea because users
1422 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1423 interpreted as invalid octal numbers. I've updated the previous comment in
1424 configure.ac, and also added a check that gives an error if 08 or 09 are
1425 used.
1427 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1428 causing partial matching to fail when the end of the subject matched \W
1429 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1431 15. There were some discrepancies between the declarations in pcre_internal.h
1432 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1433 their definitions. The declarations used "const uschar *" and the
1434 definitions used USPTR. Even though USPTR is normally defined as "const
1435 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1436 reported that: "This difference in casting confuses some C++ compilers, for
1437 example, SunCC recognizes above declarations as different functions and
1438 generates broken code for hbpcre." I have changed the declarations to use
1439 USPTR.
1441 16. GNU libtool is named differently on some systems. The autogen.sh script now
1442 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1443 (FreeBSD).
1445 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1446 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1447 comment: "Figure out how to create a longlong from a string: strtoll and
1448 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1449 instance, but it only takes 2 args instead of 3!"
1451 18. A subtle bug concerned with back references has been fixed by a change of
1452 specification, with a corresponding code fix. A pattern such as
1453 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1454 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1455 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1456 same bug. Such groups have to be quantified to be useful, or contained
1457 inside another quantified group. (If there's no repetition, the reference
1458 can never match.) The problem arises because, having left the group and
1459 moved on to the rest of the pattern, a later failure that backtracks into
1460 the group uses the captured value from the final iteration of the group
1461 rather than the correct earlier one. I have fixed this in PCRE by forcing
1462 any group that contains a reference to itself to be an atomic group; that
1463 is, there cannot be any backtracking into it once it has completed. This is
1464 similar to recursive and subroutine calls.
1467 Version 8.00 19-Oct-09
1468 ----------------------
1470 1. The table for translating pcre_compile() error codes into POSIX error codes
1471 was out-of-date, and there was no check on the pcre_compile() error code
1472 being within the table. This could lead to an OK return being given in
1473 error.
1475 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1476 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1477 in a Windows environment.
1479 3. The pcregrep --count option prints the count for each file even when it is
1480 zero, as does GNU grep. However, pcregrep was also printing all files when
1481 --files-with-matches was added. Now, when both options are given, it prints
1482 counts only for those files that have at least one match. (GNU grep just
1483 prints the file name in this circumstance, but including the count seems
1484 more useful - otherwise, why use --count?) Also ensured that the
1485 combination -clh just lists non-zero counts, with no names.
1487 4. The long form of the pcregrep -F option was incorrectly implemented as
1488 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1489 but it seems right to fix it, and I didn't think it was worth preserving
1490 the old behaviour.
1492 5. The command line items --regex=pattern and --regexp=pattern were not
1493 recognized by pcregrep, which required --regex pattern or --regexp pattern
1494 (with a space rather than an '='). The man page documented the '=' forms,
1495 which are compatible with GNU grep; these now work.
1497 6. No libpcreposix.pc file was created for pkg-config; there was just
1498 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1500 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1501 when UCP support is not needed, by modifying the Python script that
1502 generates it from Unicode data files. This should not matter if the module
1503 is correctly used as a library, but I received one complaint about 50K of
1504 unwanted data. My guess is that the person linked everything into his
1505 program rather than using a library. Anyway, it does no harm.
1507 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1508 was a minimum greater than 1 for a wide character in a possessive
1509 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1510 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1511 character. Chaos in the form of incorrect output or a compiling loop could
1512 result.
1514 9. The restrictions on what a pattern can contain when partial matching is
1515 requested for pcre_exec() have been removed. All patterns can now be
1516 partially matched by this function. In addition, if there are at least two
1517 slots in the offset vector, the offset of the earliest inspected character
1518 for the match and the offset of the end of the subject are set in them when
1519 PCRE_ERROR_PARTIAL is returned.
1521 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1522 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1523 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1524 and may be more useful for multi-segment matching.
1526 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1527 used to be given if ever the end of the subject was reached; now it is
1528 given only if matching could not proceed because another character was
1529 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1530 string "Z", which now yields "no match" instead of "partial match". In the
1531 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1532 final character ended with (*FAIL).
1534 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1535 if the pattern had a "must contain" character that was already found in the
1536 earlier partial match, unless partial matching was again requested. For
1537 example, with the pattern /dog.(body)?/, the "must contain" character is
1538 "g". If the first part-match was for the string "dog", restarting with
1539 "sbody" failed. This bug has been fixed.
1541 13. The string returned by pcre_dfa_exec() after a partial match has been
1542 changed so that it starts at the first inspected character rather than the
1543 first character of the match. This makes a difference only if the pattern
1544 starts with a lookbehind assertion or \b or \B (\K is not supported by
1545 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1546 matching functions compatible, and I think it's the right thing to do.
1548 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1549 so that the demonstration program is easily available in environments where
1550 PCRE has not been installed from source.
1552 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1553 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1554 library.
1556 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1557 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1558 is not the first non-POSIX option to be added. Clearly some people find
1559 these options useful.
1561 17. If a caller to the POSIX matching function regexec() passes a non-zero
1562 value for nmatch with a NULL value for pmatch, the value of
1563 nmatch is forced to zero.
1565 18. RunGrepTest did not have a test for the availability of the -u option of
1566 the diff command, as RunTest does. It now checks in the same way as
1567 RunTest, and also checks for the -b option.
1569 19. If an odd number of negated classes containing just a single character
1570 interposed, within parentheses, between a forward reference to a named
1571 subpattern and the definition of the subpattern, compilation crashed with
1572 an internal error, complaining that it could not find the referenced
1573 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1574 [The bug was that it was starting one character too far in when skipping
1575 over the character class, thus treating the ] as data rather than
1576 terminating the class. This meant it could skip too much.]
1578 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1579 /g option in pcretest when the pattern contains \K, which makes it possible
1580 to have an empty string match not at the start, even when the pattern is
1581 anchored. Updated pcretest and pcredemo to use this option.
1583 21. If the maximum number of capturing subpatterns in a recursion was greater
1584 than the maximum at the outer level, the higher number was returned, but
1585 with unset values at the outer level. The correct (outer level) value is
1586 now given.
1588 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1589 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1590 make it do so. The string so far is captured, making this feature
1591 compatible with Perl.
1593 23. The tests have been re-organized, adding tests 11 and 12, to make it
1594 possible to check the Perl 5.10 features against Perl 5.10.
1596 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1597 pattern matches a fixed length string. PCRE did not allow this; now it
1598 does. Neither allows recursion.
1600 25. I finally figured out how to implement a request to provide the minimum
1601 length of subject string that was needed in order to match a given pattern.
1602 (It was back references and recursion that I had previously got hung up
1603 on.) This code has now been added to pcre_study(); it finds a lower bound
1604 to the length of subject needed. It is not necessarily the greatest lower
1605 bound, but using it to avoid searching strings that are too short does give
1606 some useful speed-ups. The value is available to calling programs via
1607 pcre_fullinfo().
1609 26. While implementing 25, I discovered to my embarrassment that pcretest had
1610 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1611 study optimizations had never been tested with that matching function.
1612 Oops. What is worse, even when it was passed study data, there was a bug in
1613 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1614 were also very few tests of studied patterns with pcre_dfa_exec().
1616 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1617 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1618 on the other side of the coin, they are no longer allowed to have different
1619 names, because these cannot be distinguished in PCRE, and this has caused
1620 confusion. (This is a difference from Perl.)
1622 28. When duplicate subpattern names are present (necessarily with different
1623 numbers, as required by 27 above), and a test is made by name in a
1624 conditional pattern, either for a subpattern having been matched, or for
1625 recursion in such a pattern, all the associated numbered subpatterns are
1626 tested, and the overall condition is true if the condition is true for any
1627 one of them. This is the way Perl works, and is also more like the way
1628 testing by number works.
1631 Version 7.9 11-Apr-09
1632 ---------------------
1634 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1635 (pcretest), all targets were linked against these libraries. This included
1636 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1637 libraries. This caused unwanted dependencies to be created. This problem
1638 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1639 pcretest is linked with readline.
1641 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1642 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1643 moved outside it again, because FALSE and TRUE are already defined in AIX,
1644 but BOOL is not.
1646 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1647 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1649 4. The pcregrep documentation said spaces were inserted as well as colons (or
1650 hyphens) following file names and line numbers when outputting matching
1651 lines. This is not true; no spaces are inserted. I have also clarified the
1652 wording for the --colour (or --color) option.
1654 5. In pcregrep, when --colour was used with -o, the list of matching strings
1655 was not coloured; this is different to GNU grep, so I have changed it to be
1656 the same.
1658 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1659 each matching line was coloured. Now it goes on to look for further matches
1660 of any of the test patterns, which is the same behaviour as GNU grep.
1662 7. A pattern that could match an empty string could cause pcregrep to loop; it
1663 doesn't make sense to accept an empty string match in pcregrep, so I have
1664 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1665 seems to be how GNU grep behaves. [But see later change 40 for release
1666 8.33.]
1668 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1669 start or after a newline", because the conditional assertion was not being
1670 correctly handled. The rule now is that both the assertion and what follows
1671 in the first alternative must satisfy the test.
1673 9. If auto-callout was enabled in a pattern with a conditional group whose
1674 condition was an assertion, PCRE could crash during matching, both with
1675 pcre_exec() and pcre_dfa_exec().
1677 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1678 used for matching.
1680 11. Unicode property support in character classes was not working for
1681 characters (bytes) greater than 127 when not in UTF-8 mode.
1683 12. Added the -M command line option to pcretest.
1685 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1687 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1689 16. Added comments and documentation about mis-use of no_arg in the C++
1690 wrapper.
1692 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1693 from Martin Jerabek that uses macro names for all relevant character and
1694 string constants.
1696 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1697 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1698 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1699 these, but not everybody uses configure.
1701 19. A conditional group that had only one branch was not being correctly
1702 recognized as an item that could match an empty string. This meant that an
1703 enclosing group might also not be so recognized, causing infinite looping
1704 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1705 with the subject "ab", where knowledge that the repeated group can match
1706 nothing is needed in order to break the loop.
1708 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1709 exec(), but without supplying a callout function, matching went wrong.
1711 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1712 leak if the size of the offset vector was greater than 30. When the vector
1713 is smaller, the saved offsets during recursion go onto a local stack
1714 vector, but for larger vectors malloc() is used. It was failing to free
1715 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1716 error, in fact).
1718 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1719 heapframe that is used only when UTF-8 support is enabled. This caused no
1720 problem, but was untidy.
1722 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1723 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1724 included within another project.
1726 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1727 slightly modified by me:
1729 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1730 not building pcregrep.
1732 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1733 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1735 25. Forward references, both numeric and by name, in patterns that made use of
1736 duplicate group numbers, could behave incorrectly or give incorrect errors,
1737 because when scanning forward to find the reference group, PCRE was not
1738 taking into account the duplicate group numbers. A pattern such as
1739 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1741 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1742 the feature of a custom pointer more persuasive (as requested by a user).
1744 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1745 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1746 pre-defined.
1748 28. Added support for (*UTF8) at the start of a pattern.
1750 29. Arrange for flags added by the "release type" setting in CMake to be shown
1751 in the configuration summary.
1754 Version 7.8 05-Sep-08
1755 ---------------------
1757 1. Replaced UCP searching code with optimized version as implemented for Ad
1758 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1759 stage table and inline lookup instead of a function, giving speed ups of 2
1760 to 5 times on some simple patterns that I tested. Permission was given to
1761 distribute the MultiStage2.py script that generates the tables (it's not in
1762 the tarball, but is in the Subversion repository).
1764 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1765 scripts.
1767 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1768 a group with a zero qualifier. The result of the study could be incorrect,
1769 or the function might crash, depending on the pattern.
1771 4. Caseless matching was not working for non-ASCII characters in back
1772 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1773 It now works when Unicode Property Support is available.
1775 5. In pcretest, an escape such as \x{de} in the data was always generating
1776 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1777 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1778 truncation.
1780 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1782 7. Added two (int) casts to pcregrep when printing the difference of two
1783 pointers, in case they are 64-bit values.
1785 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1786 test 2 if it fails.
1788 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1789 and a #define of that name to empty if it is not externally set. This is to
1790 allow users of MSVC to set it if necessary.
1792 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1793 the convenience functions in the pcre_get.c source file.
1795 11. An option change at the start of a pattern that had top-level alternatives
1796 could cause overwriting and/or a crash. This command provoked a crash in
1797 some environments:
1799 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1801 This potential security problem was recorded as CVE-2008-2371.
1803 12. For a pattern where the match had to start at the beginning or immediately
1804 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1805 pcre_dfa_exec() could read past the end of the passed subject if there was
1806 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1807 pcretest so that it places the subject at the end of its malloc-ed buffer.
1809 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1810 exec() might read past the end of the data buffer in UTF-8 mode.
1812 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1813 the data contained the byte 0x85 as part of a UTF-8 character within its
1814 first line. This applied both to normal and DFA matching.
1816 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
1817 /^[^d]*?$/8 failed to match "abc".
1819 16. Added a missing copyright notice to pcrecpp_internal.h.
1821 17. Make it more clear in the documentation that values returned from
1822 pcre_exec() in ovector are byte offsets, not character counts.
1824 18. Tidied a few places to stop certain compilers from issuing warnings.
1826 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
1827 supplied by Stefan Weber. I made a further small update for 7.8 because
1828 there is a change of source arrangements: the pcre_searchfuncs.c module is
1829 replaced by pcre_ucd.c.
1832 Version 7.7 07-May-08
1833 ---------------------
1835 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
1836 a string to a long long, pretend we don't even have a long long." This is
1837 done by checking for the strtoq, strtoll, and _strtoi64 functions.
1839 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
1840 pre-7.6 versions, which defined a global no_arg variable instead of putting
1841 it in the RE class. (See also #8 below.)
1843 3. Remove a line of dead code, identified by coverity and reported by Nuno
1844 Lopes.
1846 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
1848 (1) The include/exclude patterns were being applied to the whole pathnames
1849 of files, instead of just to the final components.
1851 (2) If there was more than one level of directory, the subdirectories were
1852 skipped unless they satisfied the include/exclude conditions. This is
1853 inconsistent with GNU grep (and could even be seen as contrary to the
1854 pcregrep specification - which I improved to make it absolutely clear).
1855 The action now is always to scan all levels of directory, and just
1856 apply the include/exclude patterns to regular files.
1858 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
1859 --exclude_dir in the tests to avoid scanning .svn directories.
1861 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
1862 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
1863 doesn't support NULs in patterns.
1865 7. Added some missing "const"s to declarations of static tables in
1866 pcre_compile.c and pcre_dfa_exec.c.
1868 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
1869 caused by fix #2 above. (Subsequently also a second patch to fix the
1870 first patch. And a third patch - this was a messy problem.)
1872 9. Applied Craig's patch to remove the use of push_back().
1874 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
1875 matching function regexec().
1877 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
1878 which, however, unlike Perl's \g{...}, are subroutine calls, not back
1879 references. PCRE supports relative numbers with this syntax (I don't think
1880 Oniguruma does).
1882 12. Previously, a group with a zero repeat such as (...){0} was completely
1883 omitted from the compiled regex. However, this means that if the group
1884 was called as a subroutine from elsewhere in the pattern, things went wrong
1885 (an internal error was given). Such groups are now left in the compiled
1886 pattern, with a new opcode that causes them to be skipped at execution
1887 time.
1889 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
1890 to the way PCRE behaves:
1892 (a) A lone ] character is dis-allowed (Perl treats it as data).
1894 (b) A back reference to an unmatched subpattern matches an empty string
1895 (Perl fails the current match path).
1897 (c) A data ] in a character class must be notated as \] because if the
1898 first data character in a class is ], it defines an empty class. (In
1899 Perl it is not possible to have an empty class.) The empty class []
1900 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
1901 The negative empty class [^] matches any one character, independently
1902 of the DOTALL setting.
1904 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
1905 non-existent subpattern following a character class starting with ']' and
1906 containing () gave an internal compiling error instead of "reference to
1907 non-existent subpattern". Fortunately, when the pattern did exist, the
1908 compiled code was correct. (When scanning forwards to check for the
1909 existencd of the subpattern, it was treating the data ']' as terminating
1910 the class, so got the count wrong. When actually compiling, the reference
1911 was subsequently set up correctly.)
1913 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
1914 it was being rejected as not supported by pcre_dfa_exec(), even though
1915 other assertions are supported. I have made pcre_dfa_exec() support
1916 (*FAIL).
1918 16. The implementation of 13c above involved the invention of a new opcode,
1919 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
1920 cannot be changed at match time, I realized I could make a small
1921 improvement to matching performance by compiling OP_ALLANY instead of
1922 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
1923 on the OP_ANY path.
1925 17. Compiling pcretest on Windows with readline support failed without the
1926 following two fixes: (1) Make the unistd.h include conditional on
1927 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
1929 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
1930 ncurses library to be included for pcretest when ReadLine support is
1931 requested, but also to allow for it to be overridden. This patch came from
1932 Daniel Bergström.
1934 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
1935 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
1936 any errors with the current Unicode tables. Thanks to Peter Kankowski for
1937 spotting this.
1940 Version 7.6 28-Jan-08
1941 ---------------------
1943 1. A character class containing a very large number of characters with
1944 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
1945 overflow.
1947 2. Patch to cut out the "long long" test in pcrecpp_unittest when
1948 HAVE_LONG_LONG is not defined.
1950 3. Applied Christian Ehrlicher's patch to update the CMake build files to
1951 bring them up to date and include new features. This patch includes:
1953 - Fixed PH's badly added libz and libbz2 support.
1954 - Fixed a problem with static linking.
1955 - Added pcredemo. [But later removed - see 7 below.]
1956 - Fixed dftables problem and added an option.
1957 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
1959 - Added readline support for pcretest.
1960 - Added an listing of the option settings after cmake has run.
1962 4. A user submitted a patch to Makefile that makes it easy to create
1963 "pcre.dll" under mingw when using Configure/Make. I added stuff to
1964 Makefile.am that cause it to include this special target, without
1965 affecting anything else. Note that the same mingw target plus all
1966 the other distribution libraries and programs are now supported
1967 when configuring with CMake (see 6 below) instead of with
1968 Configure/Make.
1970 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
1971 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
1972 exported in the Windows port". It has not yet been confirmed that the patch
1973 solves the problem, but it does no harm.
1975 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
1976 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
1977 with CMake, and also correct the comment about stack recursion.
1979 7. Remove the automatic building of pcredemo from the ./configure system and
1980 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
1981 of a program that users should build themselves after PCRE is installed, so
1982 building it automatically is not really right. What is more, it gave
1983 trouble in some build environments.
1985 8. Further tidies to CMakeLists.txt from Sheri and Christian.
1988 Version 7.5 10-Jan-08
1989 ---------------------
1991 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
1992 values in parens when parsing an RE using the C++ wrapper."
1994 2. Negative specials like \S did not work in character classes in UTF-8 mode.
1995 Characters greater than 255 were excluded from the class instead of being
1996 included.
1998 3. The same bug as (2) above applied to negated POSIX classes such as
1999 [:^space:].
2001 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2002 defined or documented. It seems to have been a typo for PCRE_STATIC, so
2003 I have changed it.
2005 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
2006 first named subpattern) and a construct such as (?&a) would reference the
2007 first named subpattern whose name started with "a" (in other words, the
2008 length check was missing). Both these problems are fixed. "Subpattern name
2009 expected" is now given for (?&) (a zero-length name), and this patch also
2010 makes it give the same error for \k'' (previously it complained that that
2011 was a reference to a non-existent subpattern).
2013 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
2014 this is right because (?- can be followed by option settings as well as by
2015 digits. I have, however, made the messages clearer.
2017 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2018 than the number used in the conditional) now cause a compile-time error.
2019 This is actually not compatible with Perl, which accepts such patterns, but
2020 treats the conditional as always being FALSE (as PCRE used to), but it
2021 seems to me that giving a diagnostic is better.
2023 8. Change "alphameric" to the more common word "alphanumeric" in comments
2024 and messages.
2026 9. Fix two occurrences of "backslash" in comments that should have been
2027 "backspace".
2029 10. Remove two redundant lines of code that can never be obeyed (their function
2030 was moved elsewhere).
2032 11. The program that makes PCRE's Unicode character property table had a bug
2033 which caused it to generate incorrect table entries for sequences of
2034 characters that have the same character type, but are in different scripts.
2035 It amalgamated them into a single range, with the script of the first of
2036 them. In other words, some characters were in the wrong script. There were
2037 thirteen such cases, affecting characters in the following ranges:
2039 U+002b0 - U+002c1
2040 U+0060c - U+0060d
2041 U+0061e - U+00612
2042 U+0064b - U+0065e
2043 U+0074d - U+0076d
2044 U+01800 - U+01805
2045 U+01d00 - U+01d77
2046 U+01d9b - U+01dbf
2047 U+0200b - U+0200f
2048 U+030fc - U+030fe
2049 U+03260 - U+0327f
2050 U+0fb46 - U+0fbb1
2051 U+10450 - U+1049d
2053 12. The -o option (show only the matching part of a line) for pcregrep was not
2054 compatible with GNU grep in that, if there was more than one match in a
2055 line, it showed only the first of them. It now behaves in the same way as
2056 GNU grep.
2058 13. If the -o and -v options were combined for pcregrep, it printed a blank
2059 line for every non-matching line. GNU grep prints nothing, and pcregrep now
2060 does the same. The return code can be used to tell if there were any
2061 non-matching lines.
2063 14. Added --file-offsets and --line-offsets to pcregrep.
2065 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2066 infinitely looping recursion. The bug was that positive lookaheads were not
2067 being skipped when checking for a possible empty match (negative lookaheads
2068 and both kinds of lookbehind were skipped).
2070 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2071 inclusion of <windows.h> to before rather than after the definition of
2072 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2074 17. Specifying a possessive quantifier with a specific limit for a Unicode
2075 character property caused pcre_compile() to compile bad code, which led at
2076 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2077 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2078 caused the error; without that there was no problem.
2080 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2082 19. Added --enable-pcretest-libreadline.
2084 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2085 RE::GlobalReplace(). As a result, the number of replacements returned was
2086 double what it should be. I removed one of the increments, but Craig sent a
2087 later patch that removed the other one (the right fix) and added unit tests
2088 that check the return values (which was not done before).
2090 21. Several CMake things:
2092 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2093 the names libpcre and libpcreposix, not just pcre and pcreposix.
2095 (2) The above change means that pcretest and pcregrep are now correctly
2096 linked with the newly-built libraries, not previously installed ones.
2100 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2101 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2102 UTF-8 newline character). The key issue is that the pattern starts .*;
2103 this means that the match must be either at the beginning, or after a
2104 newline. The bug was in the code for advancing after a failed match and
2105 checking that the new position followed a newline. It was not taking
2106 account of UTF-8 characters correctly.
2108 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2109 character classes. PCRE was not treating the sequence [:...:] as a
2110 character class unless the ... were all letters. Perl, however, seems to
2111 allow any characters between [: and :], though of course it rejects as
2112 unknown any "names" that contain non-letters, because all the known class
2113 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2114 for example, whereas PCRE did not - it did not recognize a POSIX character
2115 class. This seemed a bit dangerous, so the code has been changed to be
2116 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2117 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2118 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2119 Perl does, and where it didn't before.
2121 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2122 Windows environments %n is disabled by default.
2125 Version 7.4 21-Sep-07
2126 ---------------------
2128 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2129 means that a class such as [\s] counted as "explicit reference to CR or
2130 LF". That isn't really right - the whole point of the change was to try to
2131 help when there was an actual mention of one of the two characters. So now
2132 the change happens only if \r or \n (or a literal CR or LF) character is
2133 encountered.
2135 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2136 of both had grown to the point where there were only 3 bits left.
2137 Fortunately, there was spare space in the data structure, and so I have
2138 moved the internal flags into a new 16-bit field to free up more option
2139 bits.
2141 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2142 but did not set the internal JCHANGED flag - either of these is enough to
2143 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2144 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2145 start sets both bits.
2147 4. Added options (at build time, compile time, exec time) to change \R from
2148 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2150 5. doc/pcresyntax.html was missing from the distribution.
2152 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2153 compatibility, even though it is no longer used.
2155 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2156 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2157 windows.h file is present (where different names are used). [This was
2158 reversed later after testing - see 16 below.]
2160 8. Changed all #include <config.h> to #include "config.h". There were also
2161 some further <pcre.h> cases that I changed to "pcre.h".
2163 9. When pcregrep was used with the --colour option, it missed the line ending
2164 sequence off the lines that it output.
2166 10. It was pointed out to me that arrays of string pointers cause lots of
2167 relocations when a shared library is dynamically loaded. A technique of
2168 using a single long string with a table of offsets can drastically reduce
2169 these. I have refactored PCRE in four places to do this. The result is
2170 dramatic:
2172 Originally: 290
2173 After changing UCP table: 187
2174 After changing error message table: 43
2175 After changing table of "verbs" 36
2176 After changing table of Posix names 22
2178 Thanks to the folks working on Gregex for glib for this insight.
2180 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2181 unicode-properties was also set.
2183 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2185 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2186 checked only for CRLF.
2188 14. Added casts to pcretest.c to avoid compiler warnings.
2190 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2192 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2193 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2194 entirely. This removes changes made in 7 above.
2196 17. The CMake files have been updated, and there is now more information about
2197 building with CMake in the NON-UNIX-USE document.
2200 Version 7.3 28-Aug-07
2201 ---------------------
2203 1. In the rejigging of the build system that eventually resulted in 7.1, the
2204 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2205 brackets there is not right, since it causes compilers to look for an
2206 installed pcre.h, not the version that is in the source that is being
2207 compiled (which of course may be different). I have changed it back to:
2209 #include "pcre.h"
2211 I have a vague recollection that the change was concerned with compiling in
2212 different directories, but in the new build system, that is taken care of
2213 by the VPATH setting the Makefile.
2215 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2216 when the subject happened to end in the byte 0x85 (e.g. if the last
2217 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2218 characters but of course it shouldn't be taken as a newline when it is part
2219 of another character. The bug was that, for an unlimited repeat of . in
2220 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2221 characters when looking for a newline.
2223 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2225 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2226 in debug output.
2228 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2229 long printing in the pcrecpp unittest when running under MinGW.
2231 6. ESC_K was left out of the EBCDIC table.
2233 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2234 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2235 limit also applies to "virtual nesting" when a pattern is recursive, and in
2236 this case 1000 isn't so big. I have been able to remove this limit at the
2237 expense of backing off one optimization in certain circumstances. Normally,
2238 when pcre_exec() would call its internal match() function recursively and
2239 immediately return the result unconditionally, it uses a "tail recursion"
2240 feature to save stack. However, when a subpattern that can match an empty
2241 string has an unlimited repetition quantifier, it no longer makes this
2242 optimization. That gives it a stack frame in which to save the data for
2243 checking that an empty string has been matched. Previously this was taken
2244 from the 1000-entry workspace that had been reserved. So now there is no
2245 explicit limit, but more stack is used.
2247 8. Applied Daniel's patches to solve problems with the import/export magic
2248 syntax that is required for Windows, and which was going wrong for the
2249 pcreposix and pcrecpp parts of the library. These were overlooked when this
2250 problem was solved for the main library.
2252 9. There were some crude static tests to avoid integer overflow when computing
2253 the size of patterns that contain repeated groups with explicit upper
2254 limits. As the maximum quantifier is 65535, the maximum group length was
2255 set at 30,000 so that the product of these two numbers did not overflow a
2256 32-bit integer. However, it turns out that people want to use groups that
2257 are longer than 30,000 bytes (though not repeat them that many times).
2258 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2259 made it possible to implement the integer overflow checks in a much more
2260 dynamic way, which I have now done. The artificial limitation on group
2261 length has been removed - we now have only the limit on the total length of
2262 the compiled pattern, which depends on the LINK_SIZE setting.
2264 10. Fixed a bug in the documentation for get/copy named substring when
2265 duplicate names are permitted. If none of the named substrings are set, the
2266 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2267 empty string.
2269 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2270 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2271 because the ] is interpreted as the first data character and the
2272 terminating ] is not found. PCRE has been made compatible with Perl in this
2273 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2274 cause memory overwriting.
2276 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2277 string has been matched (to stop an infinite loop). It was not recognizing
2278 a conditional subpattern that could match an empty string if that
2279 subpattern was within another subpattern. For example, it looped when
2280 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2281 condition was not nested. This bug has been fixed.
2283 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2284 past the start of the subject in the presence of bytes with the top bit
2285 set, for example "\x8aBCD".
2287 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2288 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2290 14. Optimized (?!) to (*FAIL).
2292 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2293 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2294 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2295 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2296 does: it's just the validity check that is more restrictive.
2298 16. Inserted checks for integer overflows during escape sequence (backslash)
2299 processing, and also fixed erroneous offset values for syntax errors during
2300 backslash processing.
2302 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2303 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2305 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2306 caused an overrun.
2308 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2309 something other than just ASCII characters) inside a group that had an
2310 unlimited repeat caused a loop at compile time (while checking to see
2311 whether the group could match an empty string).
2313 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2314 [\P{Any}] did so. (Error in the code for printing property names.)
2316 21. An orphan \E inside a character class could cause a crash.
2318 22. A repeated capturing bracket such as (A)? could cause a wild memory
2319 reference during compilation.
2321 23. There are several functions in pcre_compile() that scan along a compiled
2322 expression for various reasons (e.g. to see if it's fixed length for look
2323 behind). There were bugs in these functions when a repeated \p or \P was
2324 present in the pattern. These operators have additional parameters compared
2325 with \d, etc, and these were not being taken into account when moving along
2326 the compiled data. Specifically:
2328 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2329 length.
2331 (b) An item such as \pL+ within a repeated group could cause crashes or
2332 loops.
2334 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2335 "reference to non-existent subpattern" error.
2337 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2339 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2340 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2342 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2344 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2345 character were causing crashes (broken optimization).
2347 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2348 \p or \P) caused a compile-time loop.
2350 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2351 break. For example, the unstudied pattern [\r\n]A does not match the string
2352 "\r\nA" because change 7.0/46 below moves the current point on by two
2353 characters after failing to match at the start. However, the pattern \nA
2354 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2355 the same is true. There doesn't seem any very clean way out of this, but
2356 what I have chosen to do makes the common cases work: PCRE now takes note
2357 of whether there can be an explicit match for \r or \n anywhere in the
2358 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2359 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2360 pattern has explicit CR or LF references.
2362 29. Added (*CR) etc for changing newline setting at start of pattern.
2365 Version 7.2 19-Jun-07
2366 ---------------------
2368 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2369 which is apparently normally available under Windows.
2371 2. Re-jig the pcregrep tests with different newline settings in an attempt
2372 to make them independent of the local environment's newline setting.
2374 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2376 4. Some of the "internals" tests were previously cut out when the link size
2377 was not 2, because the output contained actual offsets. The recent new
2378 "Z" feature of pcretest means that these can be cut out, making the tests
2379 usable with all link sizes.
2381 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2382 stack recursion. This gives a massive performance boost under BSD, but just
2383 a small improvement under Linux. However, it saves one field in the frame
2384 in all cases.
2386 6. Added more features from the forthcoming Perl 5.10:
2388 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2389 recursion call. It refers to the nth most recently opened parentheses.
2391 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2392 to be opened parentheses.
2394 (c) Conditions that refer to capturing parentheses can be specified
2395 relatively, for example, (?(-2)... or (?(+3)...
2397 (d) \K resets the start of the current match so that everything before
2398 is not part of it.
2400 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2402 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2403 reference syntax.
2405 (g) (?| introduces a group in which the numbering of parentheses in each
2406 alternative starts with the same number.
2408 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2410 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2413 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2414 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2415 for detecting groups that can match an empty string.
2417 9. A pattern with a very large number of alternatives (more than several
2418 hundred) was running out of internal workspace during the pre-compile
2419 phase, where pcre_compile() figures out how much memory will be needed. A
2420 bit of new cunning has reduced the workspace needed for groups with
2421 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2422 workspace instead of running out of the 4096 that are available.
2424 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2426 11. Applied patch from Google to remove an optimization that didn't quite work.
2427 The report of the bug said:
2429 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2430 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2431 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2433 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2434 it matched the wrong number of bytes.
2437 Version 7.1 24-Apr-07
2438 ---------------------
2440 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2441 that is more "standard", making use of automake and other Autotools. There
2442 is some re-arrangement of the files and adjustment of comments consequent
2443 on this.
2445 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2446 for recursive directory scanning broke on some systems because the files
2447 are not scanned in any specific order and on different systems the order
2448 was different. A call to "sort" has been inserted into RunGrepTest for the
2449 approprate test as a short-term fix. In the longer term there may be an
2450 alternative.
2452 3. I had an email from Eric Raymond about problems translating some of PCRE's
2453 man pages to HTML (despite the fact that I distribute HTML pages, some
2454 people do their own conversions for various reasons). The problems
2455 concerned the use of low-level troff macros .br and .in. I have therefore
2456 removed all such uses from the man pages (some were redundant, some could
2457 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2458 HTML has been updated to handle .nf/.fi and to complain if it encounters
2459 .br or .in.
2461 4. Updated comments in configure.ac that get placed in config.h.in and also
2462 arranged for config.h to be included in the distribution, with the name
2463 config.h.generic, for the benefit of those who have to compile without
2464 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2466 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2467 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2468 makevp.bat for the current PCRE, using the additional files
2469 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2471 6. A Windows user reported a minor discrepancy with test 2, which turned out
2472 to be caused by a trailing space on an input line that had got lost in his
2473 copy. The trailing space was an accident, so I've just removed it.
2475 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2476 that is needed.
2478 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2479 as "const" (a) because they are and (b) because it helps the PHP
2480 maintainers who have recently made a script to detect big data structures
2481 in the php code that should be moved to the .rodata section. I remembered
2482 to update Builducptable as well, so it won't revert if ucptable.h is ever
2483 re-created.
2485 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2486 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2487 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2488 support is not required. This saves 1.5-2K of code, which is important in
2489 some applications.
2491 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2492 so as not to refer to the tables, even though these functions will never be
2493 called when UTF-8 support is disabled. Otherwise there are problems with a
2494 shared library.
2496 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2498 (a) It was defining its arguments as char * instead of void *.
2500 (b) It was assuming that all moves were upwards in memory; this was true
2501 a long time ago when I wrote it, but is no longer the case.
2503 The emulated memove() is provided for those environments that have neither
2504 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2505 is clearly not the case, as these two bugs were recently reported.
2507 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2508 and Detrail to create the HTML documentation, the .txt form of the man
2509 pages, and it removes trailing spaces from listed files. It also creates
2510 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2511 case, it wraps all the #defines with #ifndefs. This script should be run
2512 before "make dist".
2514 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2515 with Unicode property support.
2517 (a) For a maximizing quantifier, if the two different cases of the
2518 character were of different lengths in their UTF-8 codings (there are
2519 some cases like this - I found 11), and the matching function had to
2520 back up over a mixture of the two cases, it incorrectly assumed they
2521 were both the same length.
2523 (b) When PCRE was configured to use the heap rather than the stack for
2524 recursion during matching, it was not correctly preserving the data for
2525 the other case of a UTF-8 character when checking ahead for a match
2526 while processing a minimizing repeat. If the check also involved
2527 matching a wide character, but failed, corruption could cause an
2528 erroneous result when trying to check for a repeat of the original
2529 character.
2531 13. Some tidying changes to the testing mechanism:
2533 (a) The RunTest script now detects the internal link size and whether there
2534 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2535 values substituted by "configure". (The RunGrepTest script already did
2536 this for UTF-8.) The configure.ac script no longer substitutes the
2537 relevant variables.
2539 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2540 with length and offset values. This means that the output is different
2541 for different internal link sizes. Test 2 is skipped for link sizes
2542 other than 2 because of this, bypassing the problem. Unfortunately,
2543 there was also a test in test 3 (the locale tests) that used /B and
2544 failed for link sizes other than 2. Rather than cut the whole test out,
2545 I have added a new /Z option to pcretest that replaces the length and
2546 offset values with spaces. This is now used to make test 3 independent
2547 of link size. (Test 2 will be tidied up later.)
2549 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2550 segmentation fault instead of returning the appropriate error message.
2552 15. In multiline mode when the newline sequence was set to "any", the pattern
2553 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2554 This doesn't seem right; it now treats the CRLF combination as the line
2555 ending, and so does not match in that case. It's only a pattern such as ^$
2556 that would hit this one: something like ^ABC$ would have failed after \r
2557 and then tried again after \r\n.
2559 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2560 in an attempt to make files that differ only in their line terminators
2561 compare equal. This works on Linux.
2563 17. Under certain error circumstances pcregrep might try to free random memory
2564 as it exited. This is now fixed, thanks to valgrind.
2566 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2567 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2568 was because its rules for how to advance for /g after matching an empty
2569 string at the end of a line did not allow for this case. They now check for
2570 it specially.
2572 20. pcretest is supposed to handle patterns and data of any length, by
2573 extending its buffers when necessary. It was getting this wrong when the
2574 buffer for a data line had to be extended.
2576 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2577 CRLF as a newline sequence.
2579 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2580 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2581 I have nevertheless tidied it up.
2583 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2585 24. Added a man page for pcre-config.
2588 Version 7.0 19-Dec-06
2589 ---------------------
2591 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2592 moving to gcc 4.1.1.
2594 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2595 sys/time.h, which is documented as needed for this function. It doesn't
2596 seem to matter on Linux, but it showed up on some releases of OS X.
2598 3. It seems that there are systems where bytes whose values are greater than
2599 127 match isprint() in the "C" locale. The "C" locale should be the
2600 default when a C program starts up. In most systems, only ASCII printing
2601 characters match isprint(). This difference caused the output from pcretest
2602 to vary, making some of the tests fail. I have changed pcretest so that:
2604 (a) When it is outputting text in the compiled version of a pattern, bytes
2605 other than 32-126 are always shown as hex escapes.
2607 (b) When it is outputting text that is a matched part of a subject string,
2608 it does the same, unless a different locale has been set for the match
2609 (using the /L modifier). In this case, it uses isprint() to decide.
2611 4. Fixed a major bug that caused incorrect computation of the amount of memory
2612 required for a compiled pattern when options that changed within the
2613 pattern affected the logic of the preliminary scan that determines the
2614 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2615 that the computed length was too small. The symptoms of this bug were
2616 either the PCRE error "internal error: code overflow" from pcre_compile(),
2617 or a glibc crash with a message such as "pcretest: free(): invalid next
2618 size (fast)". Examples of patterns that provoked this bug (shown in
2619 pcretest format) are:
2621 /(?-x: )/x
2622 /(?x)(?-x: \s*#\s*)/
2623 /((?i)[\x{c0}])/8
2624 /(?i:[\x{c0}])/8
2626 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2627 is now done differently.
2629 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2630 wrapper classes; (b) implement a new function in the C++ scanner that is
2631 more efficient than the old way of doing things because it avoids levels of
2632 recursion in the regex matching; (c) add a paragraph to the documentation
2633 for the FullMatch() function.
2635 6. The escape sequence \n was being treated as whatever was defined as
2636 "newline". Not only was this contrary to the documentation, which states
2637 that \n is character 10 (hex 0A), but it also went horribly wrong when
2638 "newline" was defined as CRLF. This has been fixed.
2640 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2641 was being set to -1 for the "end of line" case (supposedly a value that no
2642 character can have). Though this value is never used (the check for end of
2643 line is "zero bytes in current character"), it caused compiler complaints.
2644 I've changed it to 0xffffffff.
2646 8. In pcre_version.c, the version string was being built by a sequence of
2647 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2648 string (as it is for production releases) called a macro with an empty
2649 argument. The C standard says the result of this is undefined. The gcc
2650 compiler treats it as an empty string (which was what was wanted) but it is
2651 reported that Visual C gives an error. The source has been hacked around to
2652 avoid this problem.
2654 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2655 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2656 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2657 of them did).
2659 10. Originally, pcretest opened its input and output without "b"; then I was
2660 told that "b" was needed in some environments, so it was added for release
2661 5.0 to both the input and output. (It makes no difference on Unix-like
2662 systems.) Later I was told that it is wrong for the input on Windows. I've
2663 now abstracted the modes into two macros, to make it easier to fiddle with
2664 them, and removed "b" from the input mode under Windows.
2666 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2668 12. Added -help and --help to pcretest as an official way of being reminded
2669 of the options.
2671 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2672 and pcrecpp.cc because they annoy compilers at high warning levels.
2674 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2676 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2677 scripts are not C programs :-) and which was not noticed because it works
2678 on Linux.
2680 16. pcretest is supposed to handle any length of pattern and data line (as one
2681 line or as a continued sequence of lines) by extending its input buffer if
2682 necessary. This feature was broken for very long pattern lines, leading to
2683 a string of junk being passed to pcre_compile() if the pattern was longer
2684 than about 50K.
2686 17. I have done a major re-factoring of the way pcre_compile() computes the
2687 amount of memory needed for a compiled pattern. Previously, there was code
2688 that made a preliminary scan of the pattern in order to do this. That was
2689 OK when PCRE was new, but as the facilities have expanded, it has become
2690 harder and harder to keep it in step with the real compile phase, and there
2691 have been a number of bugs (see for example, 4 above). I have now found a
2692 cunning way of running the real compile function in a "fake" mode that
2693 enables it to compute how much memory it would need, while actually only
2694 ever using a few hundred bytes of working memory and without too many
2695 tests of the mode. This should make future maintenance and development
2696 easier. A side effect of this work is that the limit of 200 on the nesting
2697 depth of parentheses has been removed (though this was never a serious
2698 limitation, I suspect). However, there is a downside: pcre_compile() now
2699 runs more slowly than before (30% or more, depending on the pattern). I
2700 hope this isn't a big issue. There is no effect on runtime performance.
2702 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2703 newline (only possible for the last line of a file) and it was a
2704 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2706 19. Added additional timing features to pcretest. (1) The -tm option now times
2707 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2708 separate command line item, by a number that specifies the number of
2709 repeats to use when timing. The default is 50000; this gives better
2710 precision, but takes uncomfortably long for very large patterns.
2712 20. Extended pcre_study() to be more clever in cases where a branch of a
2713 subpattern has no definite first character. For example, (a*|b*)[cd] would
2714 previously give no result from pcre_study(). Now it recognizes that the
2715 first character must be a, b, c, or d.
2717 21. There was an incorrect error "recursive call could loop indefinitely" if
2718 a subpattern (or the entire pattern) that was being tested for matching an
2719 empty string contained only one non-empty item after a nested subpattern.
2720 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2721 incorrectly, because the \d was being skipped in the check.
2723 22. The pcretest program now has a new pattern option /B and a command line
2724 option -b, which is equivalent to adding /B to every pattern. This causes
2725 it to show the compiled bytecode, without the additional information that
2726 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2727 is the same as /B/I).
2729 23. A new optimization is now able automatically to treat some sequences such
2730 as a*b as a*+b. More specifically, if something simple (such as a character
2731 or a simple class like \d) has an unlimited quantifier, and is followed by
2732 something that cannot possibly match the quantified thing, the quantifier
2733 is automatically "possessified".
2735 24. A recursive reference to a subpattern whose number was greater than 39
2736 went wrong under certain circumstances in UTF-8 mode. This bug could also
2737 have affected the operation of pcre_study().
2739 25. Realized that a little bit of performance could be had by replacing
2740 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2742 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2744 27. Possessive quantifiers such as a++ were previously implemented by turning
2745 them into atomic groups such as ($>a+). Now they have their own opcodes,
2746 which improves performance. This includes the automatically created ones
2747 from 23 above.
2749 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2750 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2751 the first matched character to be a colon. This applied both to named and
2752 numbered groups.
2754 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2756 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2757 building PCRE on Windows easier, so I have included it in the distribution.
2759 31. There is now a check in pcretest against a ridiculously large number being
2760 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2761 loop, the loop is abandoned.
2763 32. Forward references to subpatterns in conditions such as (?(2)...) where
2764 subpattern 2 is defined later cause pcre_compile() to search forwards in
2765 the pattern for the relevant set of parentheses. This search went wrong
2766 when there were unescaped parentheses in a character class, parentheses
2767 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2769 33. "Subroutine" calls and backreferences were previously restricted to
2770 referencing subpatterns earlier in the regex. This restriction has now
2771 been removed.
2773 34. Added a number of extra features that are going to be in Perl 5.10. On the
2774 whole, these are just syntactic alternatives for features that PCRE had
2775 previously implemented using the Python syntax or my own invention. The
2776 other formats are all retained for compatibility.
2778 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2779 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2780 also .NET compatible.
2782 (b) A recursion or subroutine call to a named group can now be defined as
2783 (?&name) as well as (?P>name).
2785 (c) A backreference to a named group can now be defined as \k<name> or
2786 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2787 5.10, are also .NET compatible.
2789 (d) A conditional reference to a named group can now use the syntax
2790 (?(<name>) or (?('name') as well as (?(name).
2792 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2793 groups (named and numbered) that are never evaluated inline, but can be
2794 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2795 is always false. There may be only one alternative in such a group.
2797 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2798 as the simple (?(R). The condition is true only if the most recent
2799 recursion is that of the given number or name. It does not search out
2800 through the entire recursion stack.
2802 (g) The escape \gN or \g{N} has been added, where N is a positive or
2803 negative number, specifying an absolute or relative reference.
2805 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2806 some "unreachable code" warnings.
2808 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2809 things, this adds five new scripts.
2811 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2812 There were also incompatibilities regarding the handling of \Q..\E inside
2813 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
2814 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
2816 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
2817 matches an empty string, and forcibly breaks the loop. There were bugs in
2818 this code in non-simple cases. For a pattern such as ^(a()*)* matched
2819 against aaaa the result was just "a" rather than "aaaa", for example. Two
2820 separate and independent bugs (that affected different cases) have been
2821 fixed.
2823 39. Refactored the code to abolish the use of different opcodes for small
2824 capturing bracket numbers. This is a tidy that I avoided doing when I
2825 removed the limit on the number of capturing brackets for 3.5 back in 2001.
2826 The new approach is not only tidier, it makes it possible to reduce the
2827 memory needed to fix the previous bug (38).
2829 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
2830 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
2831 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
2832 mode.
2834 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
2835 report.
2837 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
2838 copy construction and assignment in the C++ wrapper.
2840 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
2841 couple of bugs that could have given wrong results in the "--newline=crlf"
2842 case.
2844 44. Added a number of casts and did some reorganization of signed/unsigned int
2845 variables following suggestions from Dair Grant. Also renamed the variable
2846 "this" as "item" because it is a C++ keyword.
2848 45. Arranged for dftables to add
2850 #include "pcre_internal.h"
2852 to pcre_chartables.c because without it, gcc 4.x may remove the array
2853 definition from the final binary if PCRE is built into a static library and
2854 dead code stripping is activated.
2856 46. For an unanchored pattern, if a match attempt fails at the start of a
2857 newline sequence, and the newline setting is CRLF or ANY, and the next two
2858 characters are CRLF, advance by two characters instead of one.
2861 Version 6.7 04-Jul-06
2862 ---------------------
2864 1. In order to handle tests when input lines are enormously long, pcretest has
2865 been re-factored so that it automatically extends its buffers when
2866 necessary. The code is crude, but this _is_ just a test program. The
2867 default size has been increased from 32K to 50K.
2869 2. The code in pcre_study() was using the value of the re argument before
2870 testing it for NULL. (Of course, in any sensible call of the function, it
2871 won't be NULL.)
2873 3. The memmove() emulation function in pcre_internal.h, which is used on
2874 systems that lack both memmove() and bcopy() - that is, hardly ever -
2875 was missing a "static" storage class specifier.
2877 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
2878 containing an extended class (one that cannot be represented by a bitmap
2879 because it contains high-valued characters or Unicode property items, e.g.
2880 [\pZ]). Almost always one would set UTF-8 mode when processing such a
2881 pattern, but PCRE should not loop if you do not (it no longer does).
2882 [Detail: two cases were found: (a) a repeated subpattern containing an
2883 extended class; (b) a recursive reference to a subpattern that followed a
2884 previous extended class. It wasn't skipping over the extended class
2885 correctly when UTF-8 mode was not set.]
2887 5. A negated single-character class was not being recognized as fixed-length
2888 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
2889 compile error "lookbehind assertion is not fixed length".
2891 6. The RunPerlTest auxiliary script was showing an unexpected difference
2892 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
2893 write a Perl script that can interpret lines of an input file either as
2894 byte characters or as UTF-8, which is what "perltest" was being required to
2895 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
2896 can't do is switch easily at run time between having the "use utf8;" pragma
2897 or not. In the end, I fudged it by using the RunPerlTest script to insert
2898 "use utf8;" explicitly for the UTF-8 tests.
2900 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
2901 the end of the subject string, contrary to the documentation and to what
2902 Perl does. This was true of both matching functions. Now it matches only at
2903 the start of the subject and immediately after *internal* newlines.
2905 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
2906 a pointer to an int instead of a pointer to an unsigned long int. This
2907 caused problems on 64-bit systems.
2909 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
2910 instance of the 'standard' template library not being so standard".
2912 10. There was no check on the number of named subpatterns nor the maximum
2913 length of a subpattern name. The product of these values is used to compute
2914 the size of the memory block for a compiled pattern. By supplying a very
2915 long subpattern name and a large number of named subpatterns, the size
2916 computation could be caused to overflow. This is now prevented by limiting
2917 the length of names to 32 characters, and the number of named subpatterns
2918 to 10,000.
2920 11. Subpatterns that are repeated with specific counts have to be replicated in
2921 the compiled pattern. The size of memory for this was computed from the
2922 length of the subpattern and the repeat count. The latter is limited to
2923 65535, but there was no limit on the former, meaning that integer overflow
2924 could in principle occur. The compiled length of a repeated subpattern is
2925 now limited to 30,000 bytes in order to prevent this.
2927 12. Added the optional facility to have named substrings with the same name.
2929 13. Added the ability to use a named substring as a condition, using the
2930 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
2931 are numbers (not recommended). Forward references are permitted.
2933 14. Added forward references in named backreferences (if you see what I mean).
2935 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
2936 pattern could run off the end of the subject. For example, the pattern
2937 "(?s)(.{1,5})"8 did this with the subject "ab".
2939 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
2940 PCRE_CASELESS was set when matching characters that were quantified with ?
2941 or *.
2943 17. A character class other than a single negated character that had a minimum
2944 but no maximum quantifier - for example [ab]{6,} - was not handled
2945 correctly by pce_dfa_exec(). It would match only one character.
2947 18. A valid (though odd) pattern that looked like a POSIX character
2948 class but used an invalid character after [ (for example [[,abc,]]) caused
2949 pcre_compile() to give the error "Failed: internal error: code overflow" or
2950 in some cases to crash with a glibc free() error. This could even happen if
2951 the pattern terminated after [[ but there just happened to be a sequence of
2952 letters, a binary zero, and a closing ] in the memory that followed.
2954 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
2955 over the years. Originally (before any Unicode support), just the bottom 8
2956 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
2957 output from "man perlunicode" includes this:
2959 The regular expression compiler produces polymorphic opcodes. That
2960 is, the pattern adapts to the data and automatically switches to
2961 the Unicode character scheme when presented with Unicode data--or
2962 instead uses a traditional byte scheme when presented with byte
2963 data.
2965 Sadly, a wide octal escape does not cause a switch, and in a string with
2966 no other multibyte characters, these octal escapes are treated as before.
2967 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
2968 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
2969 Unicode string.
2971 I have not perpetrated such confusion in PCRE. Up till now, it took just
2972 the bottom 8 bits, as in old Perl. I have now made octal escapes with
2973 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
2974 translate to the appropriate multibyte character.
2976 29. Applied some refactoring to reduce the number of warnings from Microsoft
2977 and Borland compilers. This has included removing the fudge introduced
2978 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
2979 a warning about an unused variable.
2981 21. PCRE has not included VT (character 0x0b) in the set of whitespace
2982 characters since release 4.0, because Perl (from release 5.004) does not.
2983 [Or at least, is documented not to: some releases seem to be in conflict
2984 with the documentation.] However, when a pattern was studied with
2985 pcre_study() and all its branches started with \s, PCRE still included VT
2986 as a possible starting character. Of course, this did no harm; it just
2987 caused an unnecessary match attempt.
2989 22. Removed a now-redundant internal flag bit that recorded the fact that case
2990 dependency changed within the pattern. This was once needed for "required
2991 byte" processing, but is no longer used. This recovers a now-scarce options
2992 bit. Also moved the least significant internal flag bit to the most-
2993 significant bit of the word, which was not previously used (hangover from
2994 the days when it was an int rather than a uint) to free up another bit for
2995 the future.
2997 23. Added support for CRLF line endings as well as CR and LF. As well as the
2998 default being selectable at build time, it can now be changed at runtime
2999 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3000 specify that it is scanning data with non-default line endings.
3002 24. Changed the definition of CXXLINK to make it agree with the definition of
3003 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3005 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3006 recursions. This makes a big different to stack usage for some patterns.
3008 26. If a subpattern containing a named recursion or subroutine reference such
3009 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3010 the space required for the compiled pattern went wrong and gave too small a
3011 value. Depending on the environment, this could lead to "Failed: internal
3012 error: code overflow at offset 49" or "glibc detected double free or
3013 corruption" errors.
3015 27. Applied patches from Google (a) to support the new newline modes and (b) to
3016 advance over multibyte UTF-8 characters in GlobalReplace.
3018 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3019 difference for some implementation of PCRE in some Windows version.
3021 29. Added some extra testing facilities to pcretest:
3023 \q<number> in a data line sets the "match limit" value
3024 \Q<number> in a data line sets the "match recursion limt" value
3025 -S <number> sets the stack size, where <number> is in megabytes
3027 The -S option isn't available for Windows.
3030 Version 6.6 06-Feb-06
3031 ---------------------
3033 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3034 in pcreposix.h. I have copied the definition from pcre.h.
3036 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3037 because pcre.h is no longer a built file.
3039 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3040 not normally included in the compiled code.
3043 Version 6.5 01-Feb-06
3044 ---------------------
3046 1. When using the partial match feature with pcre_dfa_exec(), it was not
3047 anchoring the second and subsequent partial matches at the new starting
3048 point. This could lead to incorrect results. For example, with the pattern
3049 /1234/, partially matching against "123" and then "a4" gave a match.
3051 2. Changes to pcregrep:
3053 (a) All non-match returns from pcre_exec() were being treated as failures
3054 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3055 error message is output. Some extra information is given for the
3057 probably the only errors that are likely to be caused by users (by
3058 specifying a regex that has nested indefinite repeats, for instance).
3059 If there are more than 20 of these errors, pcregrep is abandoned.
3061 (b) A binary zero was treated as data while matching, but terminated the
3062 output line if it was written out. This has been fixed: binary zeroes
3063 are now no different to any other data bytes.
3065 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3066 used to set a locale for matching. The --locale=xxxx long option has
3067 been added (no short equivalent) to specify a locale explicitly on the
3068 pcregrep command, overriding the environment variables.
3070 (d) When -B was used with -n, some line numbers in the output were one less
3071 than they should have been.
3073 (e) Added the -o (--only-matching) option.
3075 (f) If -A or -C was used with -c (count only), some lines of context were
3076 accidentally printed for the final match.
3078 (g) Added the -H (--with-filename) option.
3080 (h) The combination of options -rh failed to suppress file names for files
3081 that were found from directory arguments.
3083 (i) Added the -D (--devices) and -d (--directories) options.
3085 (j) Added the -F (--fixed-strings) option.
3087 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3089 (l) Added the --colo(u)r option.
3091 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3092 is not present by default.
3094 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3095 items such as (?R) or (?1), when the recursion could match a number of
3096 alternatives. If it matched one of the alternatives, but subsequently,
3097 outside the recursion, there was a failure, the code tried to back up into
3098 the recursion. However, because of the way PCRE is implemented, this is not
3099 possible, and the result was an incorrect result from the match.
3101 In order to prevent this happening, the specification of recursion has
3102 been changed so that all such subpatterns are automatically treated as
3103 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3105 4. I had overlooked the fact that, in some locales, there are characters for
3106 which isalpha() is true but neither isupper() nor islower() are true. In
3107 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3108 and ordfeminine) are like this. This affected the treatment of \w and \W
3109 when they appeared in character classes, but not when they appeared outside
3110 a character class. The bit map for "word" characters is now created
3111 separately from the results of isalnum() instead of just taking it from the
3112 upper, lower, and digit maps. (Plus the underscore character, of course.)
3114 5. The above bug also affected the handling of POSIX character classes such as
3115 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3116 permanent tables. Instead, the bit maps for such a class were previously
3117 created as the appropriate unions of the upper, lower, and digit bitmaps.
3118 Now they are created by subtraction from the [[:word:]] class, which has
3119 its own bitmap.
3121 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3122 It is created by subtracting the vertical space characters (\x09, \x0a,
3123 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3124 subtraction was done in the overall bitmap for a character class, meaning
3125 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3126 be recognized. This bug has been fixed.
3128 7. Patches from the folks at Google:
3130 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3131 real life, but is still worth protecting against".
3133 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3134 regular expressions".
3136 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3137 have it.
3139 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3140 "configure" and the latter not, in order to fix a problem somebody had
3141 with compiling the Arg class on HP-UX.
3143 (e) Improve the error-handling of the C++ wrapper a little bit.
3145 (f) New tests for checking recursion limiting.
3147 8. The pcre_memmove() function, which is used only if the environment does not
3148 have a standard memmove() function (and is therefore rarely compiled),
3149 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3150 returning a result (though PCRE never actually uses the result).
3152 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3153 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3154 returned instead of calling malloc() with an overflowing number that would
3155 most likely cause subsequent chaos.
3157 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3159 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3160 with this option is matched, the nmatch and pmatch options of regexec() are
3161 ignored.
3163 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3164 provided in case anyone wants to the the POSIX interface with UTF-8
3165 strings.
3167 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3168 C++ linking (needed for some HP-UX environments).
3170 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3171 (unused parameter) and in the pcre_printint() function (omitted "default"
3172 switch label when the default is to do nothing).
3174 15. Added some code to make it possible, when PCRE is compiled as a C++
3175 library, to replace subject pointers for pcre_exec() with a smart pointer
3176 class, thus making it possible to process discontinuous strings.
3178 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3179 much the same function. They were added by different people who were trying
3180 to make PCRE easy to compile on non-Unix systems. It has been suggested
3181 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3182 for compiling on Windows systems. I have therefore replaced it with
3183 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3184 defaults to "extern" for C or "extern C" for C++, which works fine on
3185 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3186 SCOPE with something explicit in config.h. In addition:
3188 (a) pcreposix.h still had just "extern" instead of either of these macros;
3189 I have replaced it with PCRE_DATA_SCOPE.
3191 (b) Functions such as _pcre_xclass(), which are internal to the library,
3192 but external in the C sense, all had PCRE_EXPORT in their definitions.
3193 This is apparently wrong for the Windows case, so I have removed it.
3194 (It makes no difference on Unix-like systems.)
3196 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3197 of recursive calls to match(). This is different to MATCH_LIMIT because
3198 that limits the total number of calls to match(), not all of which increase
3199 the depth of recursion. Limiting the recursion depth limits the amount of
3200 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3201 when PCRE is compiled, and changed at run time. A patch from Google adds
3202 this functionality to the C++ interface.
3204 18. Changes to the handling of Unicode character properties:
3206 (a) Updated the table to Unicode 4.1.0.
3208 (b) Recognize characters that are not in the table as "Cn" (undefined).
3210 (c) I revised the way the table is implemented to a much improved format
3211 which includes recognition of ranges. It now supports the ranges that
3212 are defined in UnicodeData.txt, and it also amalgamates other
3213 characters into ranges. This has reduced the number of entries in the
3214 table from around 16,000 to around 3,000, thus reducing its size
3215 considerably. I realized I did not need to use a tree structure after
3216 all - a binary chop search is just as efficient. Having reduced the
3217 number of entries, I extended their size from 6 bytes to 8 bytes to
3218 allow for more data.
3220 (d) Added support for Unicode script names via properties such as \p{Han}.
3222 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3223 matching that character.
3225 20. When matching a repeated Unicode property with a minimum greater than zero,
3226 (for example \pL{2,}), PCRE could look past the end of the subject if it
3227 reached it while seeking the minimum number of characters. This could
3228 happen only if some of the characters were more than one byte long, because
3229 there is a check for at least the minimum number of bytes.
3231 21. Refactored the implementation of \p and \P so as to be more general, to
3232 allow for more different types of property in future. This has changed the
3233 compiled form incompatibly. Anybody with saved compiled patterns that use
3234 \p or \P will have to recompile them.
3236 22. Added "Any" and "L&" to the supported property types.
3238 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3239 but give a compile time error if the value is greater than 0xff.
3241 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3242 accidentally not being installed or uninstalled.
3244 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3245 made were to insert the current release number. This seemed silly, because
3246 it made things harder for people building PCRE on systems that don't run
3247 "configure". I have turned pcre.h into a distributed file, no longer built
3248 by "configure", with the version identification directly included. There is
3249 no longer a pcre.h.in file.
3251 However, this change necessitated a change to the pcre-config script as
3252 well. It is built from pcre-config.in, and one of the substitutions was the
3253 release number. I have updated configure.ac so that ./configure now finds
3254 the release number by grepping pcre.h.
3256 26. Added the ability to run the tests under valgrind.
3259 Version 6.4 05-Sep-05
3260 ---------------------
3262 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3263 "--" to be printed when multiple files were scanned, even when none of the
3264 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3265 consider it to be a bug, and have restored the previous behaviour.
3267 2. A couple of code tidies to get rid of compiler warnings.
3269 3. The pcretest program used to cheat by referring to symbols in the library
3270 whose names begin with _pcre_. These are internal symbols that are not
3271 really supposed to be visible externally, and in some environments it is
3272 possible to suppress them. The cheating is now confined to including
3273 certain files from the library's source, which is a bit cleaner.
3275 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3276 file's purpose clearer.
3278 5. Reorganized pcre_ucp_findchar().
3281 Version 6.3 15-Aug-05
3282 ---------------------
3284 1. The file libpcre.pc.in did not have general read permission in the tarball.
3286 2. There were some problems when building without C++ support:
3288 (a) If C++ support was not built, "make install" and "make test" still
3289 tried to test it.
3291 (b) There were problems when the value of CXX was explicitly set. Some
3292 changes have been made to try to fix these, and ...
3294 (c) --disable-cpp can now be used to explicitly disable C++ support.
3296 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3297 backslash in a target when C++ was disabled. This confuses some
3298 versions of "make", apparently. Using an intermediate variable solves
3299 this. (Same for CPP_LOBJ.)
3301 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3302 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3303 necessary on certain architectures.
3305 4. Added a setting of -export-symbols-regex to the link command to remove
3306 those symbols that are exported in the C sense, but actually are local
3307 within the library, and not documented. Their names all begin with
3308 "_pcre_". This is not a perfect job, because (a) we have to except some
3309 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3310 available (and never for static libraries). I have made a note to try to
3311 find a way round (a) in the future.
3314 Version 6.2 01-Aug-05
3315 ---------------------
3317 1. There was no test for integer overflow of quantifier values. A construction
3318 such as {1111111111111111} would give undefined results. What is worse, if
3319 a minimum quantifier for a parenthesized subpattern overflowed and became
3320 negative, the calculation of the memory size went wrong. This could have
3321 led to memory overwriting.
3323 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3325 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3326 operating environments where this matters.
3328 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3329 PCRE options from within the C++ wrapper.
3331 5. Named capturing subpatterns were not being correctly counted when a pattern
3332 was compiled. This caused two problems: (a) If there were more than 100
3333 such subpatterns, the calculation of the memory needed for the whole
3334 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3335 back references of the form \12, where the number was greater than 9, were
3336 not recognized as back references, even though there were sufficient
3337 previous subpatterns.
3339 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3340 versions of gcc, e.g. 2.95.4.
3343 Version 6.1 21-Jun-05
3344 ---------------------
3346 1. There was one reference to the variable "posix" in pcretest.c that was not
3347 surrounded by "#if !defined NOPOSIX".
3349 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3350 the cross-check on the old pcre_info() function, for the benefit of the
3351 cut-down version of PCRE that is currently imported into Exim.
3353 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3354 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3355 patterns that start (?i) but I don't think that matters. The use of (?i) is
3356 just an example; this all applies to the other options as well.
3358 4. Since libtool seems to echo the compile commands it is issuing, the output
3359 from "make" can be reduced a bit by putting "@" in front of each libtool
3360 compile command.
3362 5. Patch from the folks at Google for configure.in to be a bit more thorough
3363 in checking for a suitable C++ installation before trying to compile the
3364 C++ stuff. This should fix a reported problem when a compiler was present,
3365 but no suitable headers.
3367 6. The man pages all had just "PCRE" as their title. I have changed them to
3368 be the relevant file name. I have also arranged that these names are
3369 retained in the file doc/pcre.txt, which is a concatenation in text format
3370 of all the man pages except the little individual ones for each function.
3372 7. The NON-UNIX-USE file had not been updated for the different set of source
3373 files that come with release 6. I also added a few comments about the C++
3374 wrapper.
3377 Version 6.0 07-Jun-05
3378 ---------------------
3380 1. Some minor internal re-organization to help with my DFA experiments.
3382 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3383 didn't matter for the library itself when fully configured, but did matter
3384 when compiling without UCP support, or within Exim, where the ucp files are
3385 not imported.
3387 3. Refactoring of the library code to split up the various functions into
3388 different source modules. The addition of the new DFA matching code (see
3389 below) to a single monolithic source would have made it really too
3390 unwieldy, quite apart from causing all the code to be include in a
3391 statically linked application, when only some functions are used. This is
3392 relevant even without the DFA addition now that patterns can be compiled in
3393 one application and matched in another.
3395 The downside of splitting up is that there have to be some external
3396 functions and data tables that are used internally in different modules of
3397 the library but which are not part of the API. These have all had their
3398 names changed to start with "_pcre_" so that they are unlikely to clash
3399 with other external names.
3401 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3402 a different (DFA) algorithm. Although it is slower than the original
3403 function, it does have some advantages for certain types of matching
3404 problem.
3406 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3407 including restarting after a partial match.
3409 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3410 defined when compiling for Windows was sent to me. I have put it into the
3411 code, though I have no means of testing or verifying it.
3413 7. Added the pcre_refcount() auxiliary function.
3415 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3416 match before or at the first newline in the subject string. In pcretest,
3417 the /f option on a pattern can be used to set this.
3419 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3420 would behave wrongly. This has been present in PCRE since release 4.0.
3422 10. A number of changes to the pcregrep command:
3424 (a) Refactored how -x works; insert ^(...)$ instead of setting
3425 PCRE_ANCHORED and checking the length, in preparation for adding
3426 something similar for -w.
3428 (b) Added the -w (match as a word) option.
3430 (c) Refactored the way lines are read and buffered so as to have more
3431 than one at a time available.
3433 (d) Implemented a pcregrep test script.
3435 (e) Added the -M (multiline match) option. This allows patterns to match
3436 over several lines of the subject. The buffering ensures that at least
3437 8K, or the rest of the document (whichever is the shorter) is available
3438 for matching (and similarly the previous 8K for lookbehind assertions).
3440 (f) Changed the --help output so that it now says
3442 -w, --word-regex(p)
3444 instead of two lines, one with "regex" and the other with "regexp"
3445 because that confused at least one person since the short forms are the
3446 same. (This required a bit of code, as the output is generated
3447 automatically from a table. It wasn't just a text change.)
3449 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3450 option but starts with a hyphen. Could be a pattern or a path name
3451 starting with a hyphen, for instance.
3453 (h) "-" can be given as a file name to represent stdin.
3455 (i) When file names are being printed, "(standard input)" is used for
3456 the standard input, for compatibility with GNU grep. Previously
3457 "<stdin>" was used.
3459 (j) The option --label=xxx can be used to supply a name to be used for
3460 stdin when file names are being printed. There is no short form.
3462 (k) Re-factored the options decoding logic because we are going to add
3463 two more options that take data. Such options can now be given in four
3464 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3466 (l) Added the -A, -B, and -C options for requesting that lines of context
3467 around matches be printed.
3469 (m) Added the -L option to print the names of files that do not contain
3470 any matching lines, that is, the complement of -l.
3472 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3473 continue to scan other files.
3475 (o) The -s option was incorrectly implemented. For compatibility with other
3476 greps, it now suppresses the error message for a non-existent or non-
3477 accessible file (but not the return code). There is a new option called
3478 -q that suppresses the output of matching lines, which was what -s was
3479 previously doing.
3481 (p) Added --include and --exclude options to specify files for inclusion
3482 and exclusion when recursing.
3484 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3485 Hopefully, it now does.
3487 12. Missing cast in pcre_study().
3489 13. Added an "uninstall" target to the makefile.
3491 14. Replaced "extern" in the function prototypes in Makefile.in with
3492 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3493 world, but is set differently for Windows.
3495 15. Added a second compiling function called pcre_compile2(). The only
3496 difference is that it has an extra argument, which is a pointer to an
3497 integer error code. When there is a compile-time failure, this is set
3498 non-zero, in addition to the error test pointer being set to point to an
3499 error message. The new argument may be NULL if no error number is required
3500 (but then you may as well call pcre_compile(), which is now just a
3501 wrapper). This facility is provided because some applications need a
3502 numeric error indication, but it has also enabled me to tidy up the way
3503 compile-time errors are handled in the POSIX wrapper.
3505 16. Added VPATH=.libs to the makefile; this should help when building with one
3506 prefix path and installing with another. (Or so I'm told by someone who
3507 knows more about this stuff than I do.)
3509 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3510 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3511 match everything, including newlines. This is not POSIX-compatible, but
3512 somebody wanted the feature. From pcretest it can be activated by using
3513 both the P and the s flags.
3515 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3517 19. libpcre.pc was being incorrectly installed as executable.
3519 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3520 it now also looks for '\r' so that it will work unmodified on Windows.
3522 21. Added Google's contributed C++ wrapper to the distribution.
3524 22. Added some untidy missing memory free() calls in pcretest, to keep
3525 Electric Fence happy when testing.
3529 Version 5.0 13-Sep-04
3530 ---------------------
3532 1. Internal change: literal characters are no longer packed up into items
3533 containing multiple characters in a single byte-string. Each character
3534 is now matched using a separate opcode. However, there may be more than one
3535 byte in the character in UTF-8 mode.
3537 2. The pcre_callout_block structure has two new fields: pattern_position and
3538 next_item_length. These contain the offset in the pattern to the next match
3539 item, and its length, respectively.
3541 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3542 insertion of callouts before each pattern item. Added the /C option to
3543 pcretest to make use of this.
3545 4. On the advice of a Windows user, the lines
3547 #if defined(_WIN32) || defined(WIN32)
3548 _setmode( _fileno( stdout ), 0x8000 );
3549 #endif /* defined(_WIN32) || defined(WIN32) */
3551 have been added to the source of pcretest. This apparently does useful
3552 magic in relation to line terminators.
3554 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3555 for the benefit of those environments where the "b" makes a difference.
3557 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3558 to know about it. I have put a hack into configure.in that adds in code
3559 to set GCC=yes if CC=icc. This seems to end up at a point in the
3560 generated configure script that is early enough to affect the setting of
3561 compiler options, which is what is needed, but I have no means of testing
3562 whether it really works. (The user who reported this had patched the
3563 generated configure script, which of course I cannot do.)
3565 LATER: After change 22 below (new libtool files), the configure script
3566 seems to know about icc (and also ecc). Therefore, I have commented out
3567 this hack in configure.in.
3569 7. Added support for pkg-config (2 patches were sent in).
3571 8. Negated POSIX character classes that used a combination of internal tables
3572 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3573 [[:^ascii]]. Typically, they would match almost any characters. The other
3574 POSIX classes were not broken in this way.
3576 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3577 to find the match, as PCRE was deluded into thinking that the match had to
3578 start at the start point or following a newline. The same bug applied to
3579 patterns with negative forward assertions or any backward assertions
3580 preceding ".*" at the start, unless the pattern required a fixed first
3581 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3583 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3584 starting at the last subject character, bytes beyond the end of the subject
3585 string were read.
3587 11. Renamed the variable "class" as "classbits" to make life easier for C++
3588 users. (Previously there was a macro definition, but it apparently wasn't
3589 enough.)
3591 12. Added the new field "tables" to the extra data so that tables can be passed
3592 in at exec time, or the internal tables can be re-selected. This allows
3593 a compiled regex to be saved and re-used at a later time by a different
3594 program that might have everything at different addresses.
3596 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3597 -R library as well as a -L library.
3599 14. The debugging options of pcretest (-d on the command line or D on a
3600 pattern) showed incorrect output for anything following an extended class
3601 that contained multibyte characters and which was followed by a quantifier.
3603 15. Added optional support for general category Unicode character properties
3604 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3605 support. It adds about 90K to the size of the library. The meanings of the
3606 inbuilt class escapes such as \d and \s have NOT been changed.
3608 16. Updated pcredemo.c to include calls to free() to release the memory for the
3609 compiled pattern.
3611 17. The generated file chartables.c was being created in the source directory
3612 instead of in the building directory. This caused the build to fail if the
3613 source directory was different from the building directory, and was
3614 read-only.
3616 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3617 file. No doubt somebody will tell me if they don't make sense... Also added
3618 Dan Mooney's comments about building on OpenVMS.
3620 19. Added support for partial matching via the PCRE_PARTIAL option for
3621 pcre_exec() and the \P data escape in pcretest.
3623 20. Extended pcretest with 3 new pattern features:
3625 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3626 write the compiled pattern to the file whose name is "rest-of-line".
3627 This is a straight binary dump of the data, with the saved pointer to
3628 the character tables forced to be NULL. The study data, if any, is
3629 written too. After writing, pcretest reads a new pattern.
3631 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3632 compiled pattern from the given file. There must not be any
3633 occurrences of "<" in the file name (pretty unlikely); if there are,
3634 pcretest will instead treat the initial "<" as a pattern delimiter.
3635 After reading in the pattern, pcretest goes on to read data lines as
3636 usual.
3638 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3639 and 16-bit fields in a compiled pattern, to simulate a pattern that
3640 was compiled on a host of opposite endianness.
3642 21. The pcre-exec() function can now cope with patterns that were compiled on
3643 hosts of opposite endianness, with this restriction:
3645 As for any compiled expression that is saved and used later, the tables
3646 pointer field cannot be preserved; the extra_data field in the arguments
3647 to pcre_exec() should be used to pass in a tables address if a value
3648 other than the default internal tables were used at compile time.
3650 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3651 now diagnosed as an error. Previously, most of the time, a negative number
3652 would have been treated as zero, but if in addition "ovector" was passed as
3653 NULL, a crash could occur.
3655 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3656 new versions from the libtool 1.5 distribution (the last one is a copy of
3657 a file called libtool.m4). This seems to have fixed the need to patch
3658 "configure" to support Darwin 1.3 (which I used to do). However, I still
3659 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3660 workstation).
3662 24. Changed the PCRE licence to be the more standard "BSD" licence.
3665 Version 4.5 01-Dec-03
3666 ---------------------
3668 1. There has been some re-arrangement of the code for the match() function so
3669 that it can be compiled in a version that does not call itself recursively.
3670 Instead, it keeps those local variables that need separate instances for
3671 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3672 needs to "recurse". Keeping track of where control must go is done by means
3673 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3674 hide most of the details from the main code, and operates only if
3675 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3676 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3677 operating.
3679 To make it easier for callers to provide specially tailored get/free
3680 functions for this usage, two new functions, pcre_stack_malloc, and
3681 pcre_stack_free, are used. They are always called in strict stacking order,
3682 and the size of block requested is always the same.
3684 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3685 PCRE has been compiled to use the stack or the heap for recursion. The
3686 -C option of pcretest uses this to show which version is compiled.
3688 A new data escape \S, is added to pcretest; it causes the amounts of store
3689 obtained and freed by both kinds of malloc/free at match time to be added
3690 to the output.
3692 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3693 what's available on my current Linux desktop machine.
3695 3. When matching a UTF-8 string, the test for a valid string at the start has
3696 been extended. If start_offset is not zero, PCRE now checks that it points
3697 to a byte that is the start of a UTF-8 character. If not, it returns
3698 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3699 this is necessary because there may be backward assertions in the pattern.
3700 When matching the same subject several times, it may save resources to use
3701 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3703 4. The code for checking the validity of UTF-8 strings has been tightened so
3704 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3705 containing "overlong sequences".
3707 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3708 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3709 should have been "&", but it just so happened that all the cases this let
3710 through by mistake were picked up later in the function.
3712 6. I had used a variable called "isblank" - this is a C99 function, causing
3713 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3715 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3716 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3717 so the version is shown; (c) stop "make test" echoing "./RunTest".
3719 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3721 9. The prototype for memmove() for systems that don't have it was using
3722 size_t, but the inclusion of the header that defines size_t was later. I've
3723 moved the #includes for the C headers earlier to avoid this.
3725 10. Added some adjustments to the code to make it easier to compiler on certain
3726 special systems:
3728 (a) Some "const" qualifiers were missing.
3729 (b) Added the macro EXPORT before all exported functions; by default this
3730 is defined to be empty.
3731 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3732 that it reads its output file name as an argument instead of writing
3733 to the standard output and assuming this can be redirected.
3735 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3736 class containing characters with values greater than 255, PCRE compilation
3737 went into a loop.
3739 12. A recursive reference to a subpattern that was within another subpattern
3740 that had a minimum quantifier of zero caused PCRE to crash. For example,
3741 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3742 recursion. If the recursively-called subpattern itself had a zero repeat,
3743 that was OK.
3745 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3746 buffer into which it was copied (for escape processing) was still set at
3747 1024, so long lines caused crashes.
3749 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3750 "internal error: code overflow...". This applied to any character class
3751 that was followed by a possessive quantifier.
3753 15. Modified the Makefile to add libpcre.la as a prerequisite for
3754 libpcreposix.la because I was told this is needed for a parallel build to
3755 work.
3757 16. If a pattern that contained .* following optional items at the start was
3758 studied, the wrong optimizing data was generated, leading to matching
3759 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3760 matching string must start with a or b or c. The correct conclusion for
3761 this pattern is that a match can start with any character.
3764 Version 4.4 13-Aug-03
3765 ---------------------
3767 1. In UTF-8 mode, a character class containing characters with values between
3768 127 and 255 was not handled correctly if the compiled pattern was studied.
3769 In fixing this, I have also improved the studying algorithm for such
3770 classes (slightly).
3772 2. Three internal functions had redundant arguments passed to them. Removal
3773 might give a very teeny performance improvement.
3775 3. Documentation bug: the value of the capture_top field in a callout is *one
3776 more than* the number of the hightest numbered captured substring.
3778 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3779 in incorrectly linking with a previously installed version. They now link
3780 explicitly with libpcre.la.
3782 5. configure.in no longer needs to recognize Cygwin specially.
3784 6. A problem in pcre.in for Windows platforms is fixed.
3786 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3787 pcretest, it used to include the size of the study block as part of its
3788 output. Unfortunately, the structure contains a field that has a different
3789 size on different hardware architectures. This meant that the tests that
3790 showed this size failed. As the block is currently always of a fixed size,
3791 this information isn't actually particularly useful in pcretest output, so
3792 I have just removed it.
3794 8. Three pre-processor statements accidentally did not start in column 1.
3795 Sadly, there are *still* compilers around that complain, even though
3796 standard C has not required this for well over a decade. Sigh.
3798 9. In pcretest, the code for checking callouts passed small integers in the
3799 callout_data field, which is a void * field. However, some picky compilers
3800 complained about the casts involved for this on 64-bit systems. Now
3801 pcretest passes the address of the small integer instead, which should get
3802 rid of the warnings.
3804 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3805 both compile and run time, and gives an error if an invalid UTF-8 sequence
3806 is found. There is a option for disabling this check in cases where the
3807 string is known to be correct and/or the maximum performance is wanted.
3809 11. In response to a bug report, I changed one line in Makefile.in from
3811 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3812 to
3813 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
3815 to look similar to other lines, but I have no way of telling whether this
3816 is the right thing to do, as I do not use Windows. No doubt I'll get told
3817 if it's wrong...
3820 Version 4.3 21-May-03
3821 ---------------------
3823 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
3824 Makefile.
3826 2. Some refactoring to improve the quality of the code:
3828 (i) The utf8_table... variables are now declared "const".
3830 (ii) The code for \cx, which used the "case flipping" table to upper case
3831 lower case letters, now just substracts 32. This is ASCII-specific,
3832 but the whole concept of \cx is ASCII-specific, so it seems
3833 reasonable.
3835 (iii) PCRE was using its character types table to recognize decimal and
3836 hexadecimal digits in the pattern. This is silly, because it handles
3837 only 0-9, a-f, and A-F, but the character types table is locale-
3838 specific, which means strange things might happen. A private
3839 table is now used for this - though it costs 256 bytes, a table is
3840 much faster than multiple explicit tests. Of course, the standard
3841 character types table is still used for matching digits in subject
3842 strings against \d.
3844 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
3845 ending in _t are). So I've renamed it as ESC_tee.
3847 3. The first argument for regexec() in the POSIX wrapper should have been
3848 defined as "const".
3850 4. Changed pcretest to use malloc() for its buffers so that they can be
3851 Electric Fenced for debugging.
3853 5. There were several places in the code where, in UTF-8 mode, PCRE would try
3854 to read one or more bytes before the start of the subject string. Often this
3855 had no effect on PCRE's behaviour, but in some circumstances it could
3856 provoke a segmentation fault.
3858 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
3859 to try to read one or more bytes before the start of the subject string.
3861 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
3862 UTF-8 support could misbehave in various ways if the subject string
3863 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
3864 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
3865 back over UTF-8 characters.)
3868 Version 4.2 14-Apr-03
3869 ---------------------
3871 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
3873 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
3874 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
3875 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
3876 [WIN_PREFIX]: new variable, "cyg" for Cygwin
3877 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
3879 Note: automatic setting of the BUILD variables is not yet working
3880 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
3881 compile-time but not at link-time
3882 [LINK]: use for linking executables only
3883 make different versions for Windows and non-Windows
3884 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
3885 libraries
3886 [LINK_FOR_BUILD]: new variable
3887 [OBJEXT]: use throughout
3888 [EXEEXT]: use throughout
3889 <winshared>: new target
3890 <wininstall>: new target
3891 <dftables.o>: use native compiler
3892 <dftables>: use native linker
3893 <install>: handle Windows platform correctly
3894 <clean>: ditto
3895 <check>: ditto
3896 copy DLL to top builddir before testing
3898 As part of these changes, -no-undefined was removed again. This was reported
3899 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
3900 in any case.
3902 3. Some tidies to get rid of compiler warnings:
3904 . In the match_data structure, match_limit was an unsigned long int, whereas
3905 match_call_count was an int. I've made them both unsigned long ints.
3907 . In pcretest the fact that a const uschar * doesn't automatically cast to
3908 a void * provoked a warning.
3910 . Turning on some more compiler warnings threw up some "shadow" variables
3911 and a few more missing casts.
3913 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
3914 option, a class that contained a single character with a value between 128
3915 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
3917 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
3918 option, a class that contained several characters, but with at least one
3919 whose value was between 128 and 255 caused PCRE to crash.
3922 Version 4.1 12-Mar-03
3923 ---------------------
3925 1. Compiling with gcc -pedantic found a couple of places where casts were
3926 needed, and a string in dftables.c that was longer than standard compilers are
3927 required to support.
3929 2. Compiling with Sun's compiler found a few more places where the code could
3930 be tidied up in order to avoid warnings.
3932 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
3933 first of these names is deprecated in the latest Autoconf in favour of the name
3934 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
3935 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
3936 analogy I have changed it to CFLAGS_FOR_BUILD.
3938 4. Added -no-undefined to the linking command in the Makefile, because this is
3939 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
3940 linking step for the pcreposix library.
3942 5. PCRE was failing to diagnose the case of two named groups with the same
3943 name.
3945 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
3946 literal character that is needed in the subject for a match, and scans along to
3947 ensure that it is present before embarking on the full matching process. This
3948 saves time in cases of nested unlimited repeats that are never going to match.
3949 Problem: the scan can take a lot of time if the subject is very long (e.g.
3950 megabytes), thus penalizing straightforward matches. It is now done only if the
3951 amount of subject to be scanned is less than 1000 bytes.
3953 7. A lesser problem with the same optimization is that it was recording the
3954 first character of an anchored pattern as "needed", thus provoking a search
3955 right along the subject, even when the first match of the pattern was going to
3956 fail. The "needed" character is now not set for anchored patterns, unless it
3957 follows something in the pattern that is of non-fixed length. Thus, it still
3958 fulfils its original purpose of finding quick non-matches in cases of nested
3959 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
3962 Version 4.0 17-Feb-03
3963 ---------------------
3965 1. If a comment in an extended regex that started immediately after a meta-item
3966 extended to the end of string, PCRE compiled incorrect data. This could lead to
3967 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
3969 2. Moved to autoconf 2.53 and libtool 1.4.2.
3971 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
3972 the special perltest8 script is no longer needed - all the tests can be run
3973 from a single perltest script.
3975 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
3976 by \s. It has now been removed in PCRE. This means it isn't recognized as
3977 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
3978 class [:space:] *does* include VT, thereby creating a mess.
3980 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
3981 space and tab.
3983 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
3984 its new features into the main test script, reducing the number of scripts.
3986 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
3987 were backward compatible, and made the (?i) apply to the whole pattern, as if
3988 /i were given. Now it behaves more logically, and applies the option setting
3989 only to what follows. PCRE has been changed to follow suit. However, if it
3990 finds options settings right at the start of the pattern, it extracts them into
3991 the global options, as before. Thus, they show up in the info data.
3993 8. Added support for the \Q...\E escape sequence. Characters in between are
3994 treated as literals. This is slightly different from Perl in that $ and @ are
3995 also handled as literals inside the quotes. In Perl, they will cause variable
3996 interpolation. Note the following examples:
3998 Pattern PCRE matches Perl matches
4000 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
4001 \Qabc\$xyz\E abc\$xyz abc\$xyz
4002 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
4004 For compatibility with Perl, \Q...\E sequences are recognized inside character
4005 classes as well as outside them.
4007 9. Re-organized 3 code statements in pcretest to avoid "overflow in
4008 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
4009 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
4010 signed/unsigned warnings.
4012 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
4013 option for pcretest, so I've replaced it by a simple function that does just
4014 that job.
4016 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
4017 "pcregrep -".
4019 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
4020 Java package. This provides some syntactic sugar for simple cases of what my
4021 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
4022 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
4023 item, you can use this simplified notation. Note that only makes sense with
4024 greedy quantifiers. Consequently, the use of the possessive quantifier forces
4025 greediness, whatever the setting of the PCRE_UNGREEDY option.
4027 13. A change of greediness default within a pattern was not taking effect at
4028 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
4029 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
4030 was abstracted outside.
4032 14. PCRE now supports the \G assertion. It is true when the current matching
4033 position is at the start point of the match. This differs from \A when the
4034 starting offset is non-zero. Used with the /g option of pcretest (or similar
4035 code), it works in the same way as it does for Perl's /g option. If all
4036 alternatives of a regex begin with \G, the expression is anchored to the start
4037 match position, and the "anchored" flag is set in the compiled expression.
4039 15. Some bugs concerning the handling of certain option changes within patterns
4040 have been fixed. These applied to options other than (?ims). For example,
4041 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
4042 been the other way round. Some of this was related to change 7 above.
4044 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
4045 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
4046 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
4047 POSIX classes only within a class (e.g. /[[:alpha:]]/).
4049 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
4050 mode. Unlike ".", it always matches newline, whatever the setting of
4051 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
4052 assertions. Perl allows it, but it doesn't (in general) work because it can't
4053 calculate the length of the lookbehind. At least, that's the case for Perl
4054 5.8.0 - I've been told they are going to document that it doesn't work in
4055 future.
4057 18. Added an error diagnosis for escapes that PCRE does not support: these are
4058 \L, \l, \N, \P, \p, \U, \u, and \X.
4060 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4061 reading past the end of the pattern in cases such as /[abcd/.
4063 20. PCRE was getting more memory than necessary for patterns with classes that
4064 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4066 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4067 compiling PCRE for use with Virtual Pascal.
4069 22. Small fix to the Makefile to make it work properly if the build is done
4070 outside the source tree.
4072 23. Added a new extension: a condition to go with recursion. If a conditional
4073 subpattern starts with (?(R) the "true" branch is used if recursion has
4074 happened, whereas the "false" branch is used only at the top level.
4076 24. When there was a very long string of literal characters (over 255 bytes
4077 without UTF support, over 250 bytes with UTF support), the computation of how
4078 much memory was required could be incorrect, leading to segfaults or other
4079 strange effects.
4081 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4082 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4083 there was a subsequent back reference to those brackets. This meant that, for
4084 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4085 possible to check for precisely this case. All we can do is abandon the
4086 optimization if .* occurs inside capturing brackets when there are any back
4087 references whatsoever. (See below for a better fix that came later.)
4089 26. The handling of the optimization for finding the first character of a
4090 non-anchored pattern, and for finding a character that is required later in the
4091 match were failing in some cases. This didn't break the matching; it just
4092 failed to optimize when it could. The way this is done has been re-implemented.
4094 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4096 28. Added a new feature that provides some of the functionality that Perl
4097 provides with (?{...}). The facility is termed a "callout". The way it is done
4098 in PCRE is for the caller to provide an optional function, by setting
4099 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4100 global variable. By default it is unset, which disables all calling out. To get
4101 the function called, the regex must include (?C) at appropriate points. This
4102 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4103 This provides a means of identifying different callout points. When PCRE
4104 reaches such a point in the regex, if pcre_callout has been set, the external
4105 function is called. It is provided with data in a structure called
4106 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4107 matching continues; if it returns a non-zero value, the match at the current
4108 point fails. However, backtracking will occur if possible. [This was changed
4109 later and other features added - see item 49 below.]
4111 29. pcretest is upgraded to test the callout functionality. It provides a
4112 callout function that displays information. By default, it shows the start of
4113 the match and the current position in the text. There are some new data escapes
4114 to vary what happens:
4116 \C+ in addition, show current contents of captured substrings
4117 \C- do not supply a callout function
4118 \C!n return 1 when callout number n is reached
4119 \C!n!m return 1 when callout number n is reached for the mth time
4121 30. If pcregrep was called with the -l option and just a single file name, it
4122 output "<stdin>" if a match was found, instead of the file name.
4124 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4125 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4126 pcre_exec(). This saves a malloc/free per call. The default value of
4127 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4128 when configuring.
4130 32. The default maximum size of a compiled pattern is 64K. There have been a
4131 few cases of people hitting this limit. The code now uses macros to handle the
4132 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4133 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4134 configuring. Tests 2 and 5 work only with 2-byte links because they output
4135 debugging information about compiled patterns.
4137 33. Internal code re-arrangements:
4139 (a) Moved the debugging function for printing out a compiled regex into
4140 its own source file (printint.c) and used #include to pull it into
4141 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4142 separate copies.
4144 (b) Defined the list of op-code names for debugging as a macro in
4145 internal.h so that it is next to the definition of the opcodes.
4147 (c) Defined a table of op-code lengths for simpler skipping along compiled
4148 code. This is again a macro in internal.h so that it is next to the
4149 definition of the opcodes.
4151 34. Added support for recursive calls to individual subpatterns, along the
4152 lines of Robin Houston's patch (but implemented somewhat differently).
4154 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4155 allow it to read and process whole directories in Win32. This code was
4156 contributed by Lionel Fourquaux; it has not been tested by me.
4158 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4159 used to name a group. Names consist of alphanumerics and underscores, and must
4160 be unique. Back references use the syntax (?P=name) and recursive calls use
4161 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4162 numbers. The function pcre_fullinfo() can be used after compilation to extract
4163 a name/number map. There are three relevant calls:
4165 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4166 PCRE_INFO_NAMECOUNT yields the number of entries
4167 PCRE_INFO_NAMETABLE yields a pointer to the map.
4169 The map is a vector of fixed-size entries. The size of each entry depends on
4170 the length of the longest name used. The first two bytes of each entry are the
4171 group number, most significant byte first. There follows the corresponding
4172 name, zero terminated. The names are in alphabetical order.
4174 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4175 case instead of 255. Making it the same both with and without UTF-8 support
4176 means that the same test output works with both.
4178 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4179 calling malloc() with a zero argument.
4181 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4182 optimization. I've improved things by keeping a bitmap of backreferences with
4183 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4184 fact referenced, the optimization can be applied. It is unlikely that a
4185 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4186 the match to follow \n) will appear inside brackets with a number greater than
4187 31, but if it does, any back reference > 31 suppresses the optimization.
4189 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4190 of disabling numbered capturing parentheses. Any opening parenthesis that is
4191 not followed by ? behaves as if it were followed by ?: but named parentheses
4192 can still be used for capturing (and they will acquire numbers in the usual
4193 way).
4195 41. Redesigned the return codes from the match() function into yes/no/error so
4196 that errors can be passed back from deep inside the nested calls. A malloc
4197 failure while inside a recursive subpattern call now causes the
4198 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4200 42. It is now possible to set a limit on the number of times the match()
4201 function is called in a call to pcre_exec(). This facility makes it possible to
4202 limit the amount of recursion and backtracking, though not in a directly
4203 obvious way, because the match() function is used in a number of different
4204 circumstances. The count starts from zero for each position in the subject
4205 string (for non-anchored patterns). The default limit is, for compatibility, a
4206 large number, namely 10 000 000. You can change this in two ways:
4208 (a) When configuring PCRE before making, you can use --with-match-limit=n
4209 to set a default value for the compiled library.
4211 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4212 a different value is set. See 45 below.
4214 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4216 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4217 of things that can be changed at compile time. The first argument specifies
4218 what is wanted and the second points to where the information is to be placed.
4219 The current list of available information is:
4223 The output is an integer that is set to one if UTF-8 support is available;
4224 otherwise it is set to zero.
4228 The output is an integer that it set to the value of the code that is used for
4229 newline. It is either LF (10) or CR (13).
4233 The output is an integer that contains the number of bytes used for internal
4234 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4238 The output is an integer that contains the threshold above which the POSIX
4239 interface uses malloc() for output vectors. See item 31 above.
4243 The output is an unsigned integer that contains the default limit of the number
4244 of match() calls in a pcre_exec() execution. See 42 above.
4246 44. pcretest has been upgraded by the addition of the -C option. This causes it
4247 to extract all the available output from the new pcre_config() function, and to
4248 output it. The program then exits immediately.
4250 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4251 order to support additional features. One way would have been to define
4252 pcre_exec2() (for example) with extra arguments, but this would not have been
4253 extensible, and would also have required all calls to the original function to
4254 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4255 is used for passing in "extra" data from pcre_study().
4257 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4258 contains the following fields:
4260 flags a bitmap indicating which of the following fields are set
4261 study_data opaque data from pcre_study()
4262 match_limit a way of specifying a limit on match() calls for a specific
4263 call to pcre_exec()
4264 callout_data data for callouts (see 49 below)
4266 The flag bits are also defined in pcre.h, and are
4272 The pcre_study() function now returns one of these new pcre_extra blocks, with
4273 the actual study data pointed to by the study_data field, and the
4274 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4275 before. That is, this change is entirely upwards-compatible and requires no
4276 change to existing code.
4278 If you want to pass in additional data to pcre_exec(), you can either place it
4279 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4280 block.
4282 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4283 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4284 times with different match limits, until it finds the minimum value needed for
4285 pcre_exec() to complete. The value is then output. This can be instructive; for
4286 most simple matches the number is quite small, but for pathological cases it
4287 gets very large very quickly.
4289 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4290 returns the size of the data block pointed to by the study_data field in a
4291 pcre_extra block, that is, the value that was passed as the argument to
4292 pcre_malloc() when PCRE was getting memory in which to place the information
4293 created by pcre_study(). The fourth argument should point to a size_t variable.
4294 pcretest has been extended so that this information is shown after a successful
4295 pcre_study() call when information about the compiled regex is being displayed.
4297 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4298 because what follows is always an absolute path. (Later: it turns out that this
4299 is more than cosmetic for MinGW, because it doesn't like empty path
4300 components.)
4302 49. Some changes have been made to the callout feature (see 28 above):
4304 (i) A callout function now has three choices for what it returns:
4306 0 => success, carry on matching
4307 > 0 => failure at this point, but backtrack if possible
4308 < 0 => serious error, return this value from pcre_exec()
4310 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4311 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4312 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4313 use by callout functions. It will never be used by PCRE itself.
4315 (ii) The pcre_extra structure (see 45 above) has a void * field called
4316 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4317 pcre_callout_block structure has a field of the same name. The contents of
4318 the field passed in the pcre_extra structure are passed to the callout
4319 function in the corresponding field in the callout block. This makes it
4320 easier to use the same callout-containing regex from multiple threads. For
4321 testing, the pcretest program has a new data escape
4323 \C*n pass the number n (may be negative) as callout_data
4325 If the callout function in pcretest receives a non-zero value as
4326 callout_data, it returns that value.
4328 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4329 there were some redundant $(CFLAGS) in commands that are now specified as
4330 $(LINK), which already includes $(CFLAGS).
4332 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4333 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4334 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4335 one-byte characters throughout. Note that case-insensitive matching applies
4336 only to characters whose values are less than 256. PCRE doesn't support the
4337 notion of cases for higher-valued characters.
4339 (i) A character class whose characters are all within 0-255 is handled as
4340 a bit map, and the map is inverted for negative classes. Previously, a
4341 character > 255 always failed to match such a class; however it should
4342 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4344 (ii) A negated character class with a single character < 255 is coded as
4345 "not this character" (OP_NOT). This wasn't working properly when the test
4346 character was multibyte, either singly or repeated.
4348 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4349 mode, for example: \x{100}{2,3}.
4351 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4352 singly or repeated) now correctly test multibyte characters. However,
4353 PCRE doesn't recognize any characters with values greater than 255 as
4354 digits, spaces, or word characters. Such characters always match \D, \S,
4355 and \W, and never match \d, \s, or \w.
4357 (v) Classes may now contain characters and character ranges with values
4358 greater than 255. For example: [ab\x{100}-\x{400}].
4360 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4361 PCRE in UTF-8 mode.
4363 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4364 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4365 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4366 value.)
4368 53. The single man page has become too large. I have therefore split it up into
4369 a number of separate man pages. These also give rise to individual HTML pages;
4370 these are now put in a separate directory, and there is an index.html page that
4371 lists them all. Some hyperlinking between the pages has been installed.
4373 54. Added convenience functions for handling named capturing parentheses.
4375 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4376 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4377 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4378 are faulted.
4380 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4381 calling configure. These values are used when compiling the dftables.c program
4382 which is run to generate the source of the default character tables. They
4383 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4384 you will need to set these values.
4386 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4389 Version 3.9 02-Jan-02
4390 ---------------------
4392 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4394 2. If --disable-static was given, the building process failed when trying to
4395 build pcretest and pcregrep. (For some reason it was using libtool to compile
4396 them, which is not right, as they aren't part of the library.)
4399 Version 3.8 18-Dec-01
4400 ---------------------
4402 1. The experimental UTF-8 code was completely screwed up. It was packing the
4403 bytes in the wrong order. How dumb can you get?
4406 Version 3.7 29-Oct-01
4407 ---------------------
4409 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4410 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4411 this didn't happen under Solaris 8, where I normally test things.
4413 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4414 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4415 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4416 this fix an hour or so after the initial 3.7 release.)
4419 Version 3.6 23-Oct-01
4420 ---------------------
4422 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4423 offsets passed as NULL with zero offset count.
4425 2. The config.guess and config.sub files had not been updated when I moved to
4426 the latest autoconf.
4429 Version 3.5 15-Aug-01
4430 ---------------------
4432 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4433 had been forgotten.
4435 2. By using declared but undefined structures, we can avoid using "void"
4436 definitions in pcre.h while keeping the internal definitions of the structures
4437 private.
4439 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4440 user point of view, this means that both static and shared libraries are built
4441 by default, but this can be individually controlled. More of the work of
4442 handling this static/shared cases is now inside libtool instead of PCRE's make
4443 file.
4445 4. The pcretest utility is now installed along with pcregrep because it is
4446 useful for users (to test regexs) and by doing this, it automatically gets
4447 relinked by libtool. The documentation has been turned into a man page, so
4448 there are now .1, .txt, and .html versions in /doc.
4450 5. Upgrades to pcregrep:
4451 (i) Added long-form option names like gnu grep.
4452 (ii) Added --help to list all options with an explanatory phrase.
4453 (iii) Added -r, --recursive to recurse into sub-directories.
4454 (iv) Added -f, --file to read patterns from a file.
4456 6. pcre_exec() was referring to its "code" argument before testing that
4457 argument for NULL (and giving an error if it was NULL).
4459 7. Upgraded Makefile.in to allow for compiling in a different directory from
4460 the source directory.
4462 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4463 options bits, the pointer it was passed was to an int instead of to an unsigned
4464 long int. This mattered only on 64-bit systems.
4466 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4467 generated) instead of pcre.in, which it its source. Also made the same change
4468 in several of the .c files.
4470 10. A new release of gcc defines printf() as a macro, which broke pcretest
4471 because it had an ifdef in the middle of a string argument for printf(). Fixed
4472 by using separate calls to printf().
4474 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4475 script, to force use of CR or LF instead of \n in the source. On non-Unix
4476 systems, the value can be set in config.h.
4478 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4479 absolute limit. Changed the text of the error message to make this clear, and
4480 likewise updated the man page.
4482 13. The limit of 99 on the number of capturing subpatterns has been removed.
4483 The new limit is 65535, which I hope will not be a "real" limit.
4486 Version 3.4 22-Aug-00
4487 ---------------------
4489 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4491 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4494 Version 3.3 01-Aug-00
4495 ---------------------
4497 1. If an octal character was given, but the value was greater than \377, it
4498 was not getting masked to the least significant bits, as documented. This could
4499 lead to crashes in some systems.
4501 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4502 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4504 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4505 These just pass their arguments on to (pcre_free)(), but they are provided
4506 because some uses of PCRE bind it to non-C systems that can call its functions,
4507 but cannot call free() or pcre_free() directly.
4509 4. Add "make test" as a synonym for "make check". Corrected some comments in
4510 the Makefile.
4512 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4513 Makefile.
4515 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4516 command called pgrep for grepping around the active processes.
4518 7. Added the beginnings of support for UTF-8 character strings.
4520 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4521 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4522 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4523 out for the ar command.)
4526 Version 3.2 12-May-00
4527 ---------------------
4529 This is purely a bug fixing release.
4531 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4532 of ZA. This was just one example of several cases that could provoke this bug,
4533 which was introduced by change 9 of version 2.00. The code for breaking
4534 infinite loops after an iteration that matches an empty string was't working
4535 correctly.
4537 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4538 when matched against abbab (for example). After matching an empty string, it
4539 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4540 caused it to match further down the string than it should.
4542 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4543 was there because it doesn't seem to be needed, and it causes trouble on some
4544 systems, as it is not a Standard C header. It has been removed.
4546 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4547 were reported on the Macintosh. The changes were from
4549 while ((c = *(++ptr)) != 0 && c != '\n');
4550 to
4551 while ((c = *(++ptr)) != 0 && c != '\n') ;
4553 Totally extraordinary, but if that's what it takes...
4555 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4556 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4557 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4558 assumes the way PCRE uses memmove() (always moving upwards).
4560 6. PCRE is being used in one environment where strchr() is not available. There
4561 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4562 faster code anyway.
4565 Version 3.1 09-Feb-00
4566 ---------------------
4568 The only change in this release is the fixing of some bugs in Makefile.in for
4569 the "install" target:
4571 (1) It was failing to install pcreposix.h.
4573 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4576 Version 3.0 01-Feb-00
4577 ---------------------
4579 1. Add support for the /+ modifier to perltest (to output $` like it does in
4580 pcretest).
4582 2. Add support for the /g modifier to perltest.
4584 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4585 matches null strings.
4587 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4588 pattern. Perl treats empty patterns specially - it reuses the most recent
4589 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4590 effect.
4592 5. The POSIX interface was broken in that it was just handing over the POSIX
4593 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4594 required a bigger vector, with some working space on the end. This means that
4595 the POSIX wrapper now has to get and free some memory, and copy the results.
4597 6. Added some simple autoconf support, placing the test data and the
4598 documentation in separate directories, re-organizing some of the
4599 information files, and making it build pcre-config (a GNU standard). Also added
4600 libtool support for building PCRE as a shared library, which is now the
4601 default.
4603 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4604 09 are not valid octal constants. Single digits will be used for minor values
4605 less than 10.
4607 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4608 existing programs that set these in the POSIX interface can use PCRE without
4609 modification.
4611 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4612 return all that pcre_info() returns, plus additional data. The pcre_info()
4613 function is retained for compatibility, but is considered to be obsolete.
4615 10. Added experimental recursion feature (?R) to handle one common case that
4616 Perl 5.6 will be able to do with (?p{...}).
4618 11. Added support for POSIX character classes like [:alpha:], which Perl is
4619 adopting.
4622 Version 2.08 31-Aug-99
4623 ----------------------
4625 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4626 trying to match at the startoffset position, but instead was moving forward to
4627 the next newline as if a previous match had failed.
4629 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4630 and could get into a loop if a null string was matched other than at the start
4631 of the subject.
4633 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4634 be distinguished at compile time, and for completeness also added PCRE_DATE.
4636 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4637 in GnuWin32 environments.
4640 Version 2.07 29-Jul-99
4641 ----------------------
4643 1. The documentation is now supplied in plain text form and HTML as well as in
4644 the form of man page sources.
4646 2. C++ compilers don't like assigning (void *) values to other pointer types.
4647 In particular this affects malloc(). Although there is no problem in Standard
4648 C, I've put in casts to keep C++ compilers happy.
4650 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4651 should be (const char *).
4653 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4654 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4655 However, I haven't made this a standard facility. The documentation doesn't
4656 mention it, and the Makefile doesn't support it.
4658 5. The Makefile now contains an "install" target, with editable destinations at
4659 the top of the file. The pcretest program is not installed.
4661 6. pgrep -V now gives the PCRE version number and date.
4663 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4664 causing the entire string to be ignored, instead of just the last character.
4666 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4667 non-matching string, it can take a very, very long time, even for strings of
4668 quite modest length, because of the nested recursion. PCRE now does better in
4669 some of these cases. It does this by remembering the last required literal
4670 character in the pattern, and pre-searching the subject to ensure it is present
4671 before running the real match. In other words, it applies a heuristic to detect
4672 some types of certain failure quickly, and in the above example, if presented
4673 with a string that has no trailing " it gives "no match" very quickly.
4675 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4676 other alternatives are tried instead.
4679 Version 2.06 09-Jun-99
4680 ----------------------
4682 1. Change pcretest's output for amount of store used to show just the code
4683 space, because the remainder (the data block) varies in size between 32-bit and
4684 64-bit systems.
4686 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4687 start matching at. This allows lookbehinds to work when searching for multiple
4688 occurrences in a string.
4690 3. Added additional options to pcretest for testing multiple occurrences:
4692 /+ outputs the rest of the string that follows a match
4693 /g loops for multiple occurrences, using the new startoffset argument
4694 /G loops for multiple occurrences by passing an incremented pointer
4696 4. PCRE wasn't doing the "first character" optimization for patterns starting
4697 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4698 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4699 the letter 'x'. On long subject strings, this gives a significant speed-up.
4702 Version 2.05 21-Apr-99
4703 ----------------------
4705 1. Changed the type of magic_number from int to long int so that it works
4706 properly on 16-bit systems.
4708 2. Fixed a bug which caused patterns starting with .* not to work correctly
4709 when the subject string contained newline characters. PCRE was assuming
4710 anchoring for such patterns in all cases, which is not correct because .* will
4711 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4712 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4713 must be retried after every newline in the subject.
4716 Version 2.04 18-Feb-99
4717 ----------------------
4719 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4720 computation of the store needed to hold the pattern was incorrect (too large).
4721 If such patterns were nested a few deep, this could multiply and become a real
4722 problem.
4724 2. Added /M option to pcretest to show the memory requirement of a specific
4725 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4727 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4728 compiled in such a way that the backtracking after subsequent failure was
4729 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4730 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4733 Version 2.03 02-Feb-99
4734 ----------------------
4736 1. Fixed typo and small mistake in man page.
4738 2. Added 4th condition (GPL supersedes if conflict) and created separate
4739 LICENCE file containing the conditions.
4741 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4742 Perl, that is the internal \ allows the delimiter to be included in the
4743 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4744 the final delimiter, add \ to the end of the pattern (to test the error).
4746 4. Added the convenience functions for extracting substrings after a successful
4747 match. Updated pcretest to make it able to test these functions.
4750 Version 2.02 14-Jan-99
4751 ----------------------
4753 1. Initialized the working variables associated with each extraction so that
4754 their saving and restoring doesn't refer to uninitialized store.
4756 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4757 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4758 fix the problem.
4760 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4761 calls, and wasn't printing the correct value for compiling calls. Increased the
4762 default value of LOOPREPEAT, and the number of significant figures in the
4763 times.
4765 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4767 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4768 a building problem on Windows NT with a FAT file system.
4771 Version 2.01 21-Oct-98
4772 ----------------------
4774 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4775 to character tables built by pcre_maketables() in the current locale. If NULL
4776 is passed, the default tables are used.
4779 Version 2.00 24-Sep-98
4780 ----------------------
4782 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4783 it any more.
4785 2. Allow quantification of (?>) groups, and make it work correctly.
4787 3. The first character computation wasn't working for (?>) groups.
4789 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4790 end of the subject) and add 5.005's \z, which really does match only at the
4791 very end of the subject.
4793 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4795 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4796 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4797 localized options. All options to pcre_study() were also removed.
4799 7. Add other new features from 5.005:
4801 $(?<= positive lookbehind
4802 $(?<! negative lookbehind
4803 (?imsx-imsx) added the unsetting capability
4804 such a setting is global if at outer level; local otherwise
4805 (?imsx-imsx:) non-capturing groups with option setting
4806 (?(cond)re|re) conditional pattern matching
4808 A backreference to itself in a repeated group matches the previous
4809 captured string.
4811 8. General tidying up of studying (both automatic and via "study")
4812 consequential on the addition of new assertions.
4814 9. As in 5.005, unlimited repeated groups that could match an empty substring
4815 are no longer faulted at compile time. Instead, the loop is forcibly broken at
4816 runtime if any iteration does actually match an empty substring.
4818 10. Include the RunTest script in the distribution.
4820 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
4821 discrepancies, some of which were old and were also with respect to 5.004. They
4822 have now been fixed.
4825 Version 1.09 28-Apr-98
4826 ----------------------
4828 1. A negated single character class followed by a quantifier with a minimum
4829 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
4830 program crashes, or just wrong answers. This did not apply to negated classes
4831 containing more than one character, or to minima other than one.
4834 Version 1.08 27-Mar-98
4835 ----------------------
4837 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
4839 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
4840 latter must appear before anything that relies on it in the pattern.
4843 Version 1.07 16-Feb-98
4844 ----------------------
4846 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
4847 repeat of a potentially empty string).
4850 Version 1.06 23-Jan-98
4851 ----------------------
4853 1. Added Markus Oberhumer's little patches for C++.
4855 2. Literal strings longer than 255 characters were broken.
4858 Version 1.05 23-Dec-97
4859 ----------------------
4861 1. Negated character classes containing more than one character were failing if
4862 PCRE_CASELESS was set at run time.
4865 Version 1.04 19-Dec-97
4866 ----------------------
4868 1. Corrected the man page, where some "const" qualifiers had been omitted.
4870 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
4871 input syntax.
4873 3. Fixed memory leak which occurred when a regex with back references was
4874 matched with an offsets vector that wasn't big enough. The temporary memory
4875 that is used in this case wasn't being freed if the match failed.
4877 4. Tidied pcretest to ensure it frees memory that it gets.
4879 5. Temporary memory was being obtained in the case where the passed offsets
4880 vector was exactly big enough.
4882 6. Corrected definition of offsetof() from change 5 below.
4884 7. I had screwed up change 6 below and broken the rules for the use of
4885 setjmp(). Now fixed.
4888 Version 1.03 18-Dec-97
4889 ----------------------
4891 1. A erroneous regex with a missing opening parenthesis was correctly
4892 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
4893 on some systems.
4895 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
4896 it was reported that one broken compiler failed on the former because "code" is
4897 also an independent variable.
4899 3. The erroneous regex a[]b caused an array overrun reference.
4901 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
4902 fail on data ending with that character. (It was going on too far, and checking
4903 the next character, typically a binary zero.) This was specific to the
4904 optimized code for single-character negative classes.
4906 5. Added a contributed patch from the TIN world which does the following:
4908 + Add an undef for memmove, in case the the system defines a macro for it.
4910 + Add a definition of offsetof(), in case there isn't one. (I don't know
4911 the reason behind this - offsetof() is part of the ANSI standard - but
4912 it does no harm).
4914 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
4915 most of the places where whitespace preceded '#'. I have given up and
4916 allowed the remaining 2 cases to be at the margin.
4918 + Rename some variables in pcre to eliminate shadowing. This seems very
4919 pedantic, but does no harm, of course.
4921 6. Moved the call to setjmp() into its own function, to get rid of warnings
4922 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
4924 7. Constructs such as \d{8,} were compiling into the equivalent of
4925 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
4926 outcome, but in this particular case used more store than had been allocated,
4927 which caused the bug to be discovered because it threw up an internal error.
4929 8. The debugging code in both pcre and pcretest for outputting the compiled
4930 form of a regex was going wrong in the case of back references followed by
4931 curly-bracketed repeats.
4934 Version 1.02 12-Dec-97
4935 ----------------------
4937 1. Typos in pcre.3 and comments in the source fixed.
4939 2. Applied a contributed patch to get rid of places where it used to remove
4940 'const' from variables, and fixed some signed/unsigned and uninitialized
4941 variable warnings.
4943 3. Added the "runtest" target to Makefile.
4945 4. Set default compiler flag to -O2 rather than just -O.
4948 Version 1.01 19-Nov-97
4949 ----------------------
4951 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
4952 like /([ab]*)*/, that is, for classes with more than one character in them.
4954 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
4955 as /((?>a*))*/ (a PCRE_EXTRA facility).
4958 Version 1.00 18-Nov-97
4959 ----------------------
4961 1. Added compile-time macros to support systems such as SunOS4 which don't have
4962 memmove() or strerror() but have other things that can be used instead.
4964 2. Arranged that "make clean" removes the executables.
4967 Version 0.99 27-Oct-97
4968 ----------------------
4970 1. Fixed bug in code for optimizing classes with only one character. It was
4971 initializing a 32-byte map regardless, which could cause it to run off the end
4972 of the memory it had got.
4974 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
4977 Version 0.98 22-Oct-97
4978 ----------------------
4980 1. Fixed bug in code for handling temporary memory usage when there are more
4981 back references than supplied space in the ovector. This could cause segfaults.
4984 Version 0.97 21-Oct-97
4985 ----------------------
4987 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
4989 2. Optimized negated single characters not to use a bit map.
4991 3. Brought error texts together as macro definitions; clarified some of them;
4992 fixed one that was wrong - it said "range out of order" when it meant "invalid
4993 escape sequence".
4995 4. Changed some char * arguments to const char *.
4997 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
4999 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
5000 pcretest.
5003 Version 0.96 16-Oct-97
5004 ----------------------
5006 1. Added a simple "pgrep" utility to the distribution.
5008 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
5009 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
5010 where "ddd" means "one or more decimal digits".
5012 3. Fixed serious bug. If a pattern had a back reference, but the call to
5013 pcre_exec() didn't supply a large enough ovector to record the related
5014 identifying subpattern, the match always failed. PCRE now remembers the number
5015 of the largest back reference, and gets some temporary memory in which to save
5016 the offsets during matching if necessary, in order to ensure that
5017 backreferences always work.
5019 4. Increased the compatibility with Perl in a number of ways:
5021 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
5022 to request this handling. The option can be set at compile or exec time.
5024 (b) $ matches before a terminating newline by default; an option
5025 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
5026 mode). The option can be set at compile or exec time.
5028 (c) The handling of \ followed by a digit other than 0 is now supposed to be
5029 the same as Perl's. If the decimal number it represents is less than 10
5030 or there aren't that many previous left capturing parentheses, an octal
5031 escape is read. Inside a character class, it's always an octal escape,
5032 even if it is a single digit.
5034 (d) An escaped but undefined alphabetic character is taken as a literal,
5035 unless PCRE_EXTRA is set. Currently this just reserves the remaining
5036 escapes.
5038 (e) {0} is now permitted. (The previous item is removed from the compiled
5039 pattern).
5041 5. Changed all the names of code files so that the basic parts are no longer
5042 than 10 characters, and abolished the teeny "globals.c" file.
5044 6. Changed the handling of character classes; they are now done with a 32-byte
5045 bit map always.
5047 7. Added the -d and /D options to pcretest to make it possible to look at the
5048 internals of compilation without having to recompile pcre.
5051 Version 0.95 23-Sep-97
5052 ----------------------
5054 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
5055 \x20 at the start of a run of normal characters. These were being treated as
5056 real characters, instead of the source characters being re-checked.
5059 Version 0.94 18-Sep-97
5060 ----------------------
5062 1. The functions are now thread-safe, with the caveat that the global variables
5063 containing pointers to malloc() and free() or alternative functions are the
5064 same for all threads.
5066 2. Get pcre_study() to generate a bitmap of initial characters for non-
5067 anchored patterns when this is possible, and use it if passed to pcre_exec().
5070 Version 0.93 15-Sep-97
5071 ----------------------
5073 1. /(b)|(:+)/ was computing an incorrect first character.
5075 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5076 but not actually doing anything yet.
5078 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5079 as Perl does (e.g. [-az] or [az-]).
5081 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5082 all possible positions.
5084 5. Split up into different modules to avoid including unneeded functions in a
5085 compiled binary. However, compile and exec are still in one module. The "study"
5086 function is split off.
5088 6. The character tables are now in a separate module whose source is generated
5089 by an auxiliary program - but can then be edited by hand if required. There are
5090 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5091 toupper() in the code.
5093 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5094 make them global. Abolish the function for setting them, as the caller can now
5095 set them directly.
5098 Version 0.92 11-Sep-97
5099 ----------------------
5101 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5102 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5104 2. Caseless matching was not working in character classes if the characters in
5105 the pattern were in upper case.
5107 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5109 4. Make PCRE_ANCHORED public and accept as a compile option.
5111 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5112 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5113 pass them.
5115 6. Give an error if bad option bits passed at compile or run time.
5117 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5118 pcretest to cause it to pass that flag.
5120 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5121 options, and the first character, if set.
5123 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5126 Version 0.91 10-Sep-97
5127 ----------------------
5129 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5130 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5132 2. PCRE was looping on encountering an indefinitely repeated back reference to
5133 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5134 Perl does - treats the match as successful.
5136 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5