ViewVC logotype

Contents of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1394 - (show annotations)
Sat Nov 9 09:17:20 2013 UTC (5 years, 11 months ago) by ph10
File size: 256528 byte(s)
Require group names to start with a non-digit.
1 ChangeLog for PCRE
2 ------------------
4 Version 8.34 xx-xxxx-201x
5 --------------------------
7 1. Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
8 executable memory. Patch inspired by Carsten Klein.
10 2. ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
11 this macro is never tested and has no effect, because the work to support
12 coverage involves only compiling and linking options and special targets in
13 the Makefile. The comment in config.h implied that defining the macro would
14 enable coverage support, which is totally false. There was also support for
15 setting this macro in the CMake files (my fault, I just copied it from
16 configure). SUPPORT_GCOV has now been removed.
18 3. Make a small performance improvement in strlen16() and strlen32() in
19 pcretest.
21 4. Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
22 detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
23 these cases). There was also one in pcretest.c.
25 5. Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
27 6. In UTF mode, the code for checking whether a group could match an empty
28 string (which is used for indefinitely repeated groups to allow for
29 breaking an infinite loop) was broken when the group contained a repeated
30 negated single-character class with a character that occupied more than one
31 data item and had a minimum repetition of zero (for example, [^\x{100}]* in
32 UTF-8 mode). The effect was undefined: the group might or might not be
33 deemed as matching an empty string, or the program might have crashed.
35 7. The code for checking whether a group could match an empty string was not
36 recognizing that \h, \H, \v, \V, and \R must match a character.
38 8. Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
39 an empty string. If it can, pcretest shows this in its information output.
41 9. Fixed two related bugs that applied to Unicode extended grapheme clusters
42 that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
43 matched by pcre_exec() without using JIT:
45 (a) If the rest of the pattern did not match after a maximal run of
46 grapheme clusters, the code for backing up to try with fewer of them
47 did not always back up over a full grapheme when characters that do not
48 have the modifier quality were involved, e.g. Hangul syllables.
50 (b) If the match point in a subject started with modifier character, and
51 there was no match, the code could incorrectly back up beyond the match
52 point, and potentially beyond the first character in the subject,
53 leading to a segfault or an incorrect match result.
55 10. A conditional group with an assertion condition could lead to PCRE
56 recording an incorrect first data item for a match if no other first data
57 item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
58 first data item, and therefore matched "ca" after "c" instead of at the
59 start.
61 11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
62 bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
64 12. The source of pcregrep now includes z/OS-specific code so that it can be
65 compiled for z/OS as part of the special z/OS distribution.
67 13. Added the -T and -TM options to pcretest.
69 14. The code in pcre_compile.c for creating the table of named capturing groups
70 has been refactored. Instead of creating the table dynamically during the
71 actual compiling pass, the information is remembered during the pre-compile
72 pass (on the stack unless there are more than 20 named groups, in which
73 case malloc() is used) and the whole table is created before the actual
74 compile happens. This has simplified the code (it is now nearly 150 lines
75 shorter) and prepared the way for better handling of references to groups
76 with duplicate names.
78 15. A back reference to a named subpattern when there is more than one of the
79 same name now checks them in the order in which they appear in the pattern.
80 The first one that is set is used for the reference. Previously only the
81 first one was inspected. This change makes PCRE more compatible with Perl.
83 16. Unicode character properties were updated from Unicode 6.3.0.
85 17. The compile-time code for auto-possessification has been refactored, based
86 on a patch by Zoltan Herczeg. It now happens after instead of during
87 compilation. The code is cleaner, and more cases are handled. The option
88 PCRE_NO_AUTO_POSSESSIFY is added for testing purposes, and the -O and /O
89 options in pcretest are provided to set it.
91 18. The character VT has been added to the set of characters that match \s and
92 are generally treated as white space, following this same change in Perl
93 5.18. There is now no difference between "Perl space" and "POSIX space".
95 19. The code for checking named groups as conditions, either for being set or
96 for being recursed, has been refactored (this is related to 14 and 15
97 above). Processing unduplicated named groups should now be as fast at
98 numerical groups, and processing duplicated groups should be faster than
99 before.
101 20. Two patches to the CMake build system, by Alexander Barkov:
103 (1) Replace the "source" command by "." in CMakeLists.txt because
104 "source" is a bash-ism.
106 (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
107 without these the CMake build does not work on Solaris.
109 21. Perl has changed its handling of \8 and \9. If there is no previously
110 encountered capturing group of those numbers, they are treated as the
111 literal characters 8 and 9 instead of a binary zero followed by the
112 literals. PCRE now does the same.
114 22. Following Perl, added \o{} to specify codepoints in octal, making it
115 possible to specify values greater than 0777 and also making them
116 unambiguous.
118 23. Perl now gives an error for missing closing braces after \x{... instead of
119 treating the string as literal. PCRE now does the same.
121 24. RunTest used to grumble if an inappropriate test was selected explicitly,
122 but just skip it when running all tests. This make it awkward to run ranges
123 of tests when one of them was inappropriate. Now it just skips any
124 inappropriate tests, as it always did when running all tests.
126 25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
127 character types such as \d or \w, too many callouts were inserted, and the
128 data that they returned was rubbish.
130 26. In UCP mode, \s was not matching two of the characters that Perl matches,
131 namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
132 were matched by \h. The code has now been refactored so that the lists of
133 the horizontal and vertical whitespace characters used for \h and \v (which
134 are defined only in one place) are now also used for \s.
136 27. Add JIT support for the 64 bit TileGX architecture.
137 Patch by Jiong Wang (Tilera Corporation).
139 28. Possessive quantifiers for classes (both explicit and automatically
140 generated) now use special opcodes instead of wrapping in ONCE brackets.
142 29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
143 (because it's meaningless), this was not happening when PCRE_CASELESS was
144 set. Not wrong, but inefficient.
146 30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
147 properties for \w, \d, etc) is present in a test regex. Otherwise if the
148 test contains no characters greater than 255, Perl doesn't realise it
149 should be using Unicode semantics.
151 31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
152 [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
153 does in Unicode mode.
155 32. Added the "forbid" facility to pcretest so that putting tests into the
156 wrong test files can sometimes be quickly detected.
158 33. There is now a limit (default 250) on the depth of nesting of parentheses.
159 This limit is imposed to control the amount of system stack used at compile
160 time. It can be changed at build time by --with-parens-nest-limit=xxx or
161 the equivalent in CMake.
163 34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
164 errors. Perl warns for these when in warning mode, but PCRE has no facility
165 for giving warnings.
167 35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
168 However, this was not working for (?!) because it is optimized to (*FAIL),
169 for which PCRE does not allow quantifiers. The optimization is now disabled
170 when a quantifier follows (?!). I can't see any use for this, but it makes
171 things uniform.
173 36. Perl no longer allows group names to start with digits, so I have made this
174 change also in PCRE. It simplifies the code a bit.
177 Version 8.33 28-May-2013
178 ------------------------
180 1. Added 'U' to some constants that are compared to unsigned integers, to
181 avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
182 variables that are added to signed variables, to ensure the result is
183 signed and can be negated.
185 2. Applied patch by Daniel Richard G for quashing MSVC warnings to the
186 CMake config files.
188 3. Revise the creation of config.h.generic so that all boolean macros are
189 #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
190 overriding via -D on the command line possible.
192 4. Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
193 to unsigned int is reported to make a quite noticeable speed difference in
194 a specific Windows environment. Testing on Linux did also appear to show
195 some benefit (and it is clearly not harmful). Also fixed the definition of
196 Xop which should be unsigned.
198 5. Related to (4), changing the definition of the intermediate variable cc
199 in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
200 improvements.
202 6. Fix forward search in JIT when link size is 3 or greater. Also removed some
203 unnecessary spaces.
205 7. Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
206 and later.
208 8. Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
210 9. Optimizing fast_forward_start_bits in JIT.
212 10. Adding support for callouts in JIT, and fixing some issues revealed
213 during this work. Namely:
215 (a) Unoptimized capturing brackets incorrectly reset on backtrack.
217 (b) Minimum length was not checked before the matching is started.
219 11. The value of capture_last that is passed to callouts was incorrect in some
220 cases when there was a capture on one path that was subsequently abandoned
221 after a backtrack. Also, the capture_last value is now reset after a
222 recursion, since all captures are also reset in this case.
224 12. The interpreter no longer returns the "too many substrings" error in the
225 case when an overflowing capture is in a branch that is subsequently
226 abandoned after a backtrack.
228 13. In the pathological case when an offset vector of size 2 is used, pcretest
229 now prints out the matched string after a yield of 0 or 1.
231 14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
232 Only supported by the JIT compiler at the moment.
234 15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
236 16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
237 offset of the starting point of the matching process, provided the offsets
238 vector is large enough.
240 17. The \A escape now records a lookbehind value of 1, though its execution
241 does not actually inspect the previous character. This is to ensure that,
242 in partial multi-segment matching, at least one character from the old
243 segment is retained when a new segment is processed. Otherwise, if there
244 are no lookbehinds in the pattern, \A might match incorrectly at the start
245 of a new segment.
247 18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
249 19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
250 gives some modest performance improvement in 8-bit mode.
252 20. Added the PCRE-specific property \p{Xuc} for matching characters that can
253 be expressed in certain programming languages using Universal Character
254 Names.
256 21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
257 which points out that "non characters" are not "characters that may not
258 appear in Unicode strings" but rather "characters that are reserved for
259 internal use and have only local meaning".
261 22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
262 there was a conditional group that depended on an assertion, if the
263 assertion was false, the callout that immediately followed the alternation
264 in the condition was skipped when pcre_exec() was used for matching.
266 23. Allow an explicit callout to be inserted before an assertion that is the
267 condition for a conditional group, for compatibility with automatic
268 callouts, which always insert a callout at this point.
270 24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
271 confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
273 25. (*PRUNE) is now supported by the JIT compiler.
275 26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
277 27. Fix the case where there are two or more SKIPs with arguments that may be
278 ignored.
280 28. (*SKIP) is now supported by the JIT compiler.
282 29. (*THEN) is now supported by the JIT compiler.
284 30. Update RunTest with additional test selector options.
286 31. The way PCRE handles backtracking verbs has been changed in two ways.
288 (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
289 SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
290 backtracking. In some cases this makes it more Perl-compatible, but Perl's
291 rather obscure rules do not always do the same thing.
293 (2) Previously, backtracking verbs were confined within assertions. This is
294 no longer the case for positive assertions, except for (*ACCEPT). Again,
295 this sometimes improves Perl compatibility, and sometimes does not.
297 32. A number of tests that were in test 2 because Perl did things differently
298 have been moved to test 1, because either Perl or PCRE has changed, and
299 these tests are now compatible.
301 32. Backtracking control verbs are now handled in the same way in JIT and
302 interpreter.
304 33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
305 contained a forward subroutine reference caused a compile error.
307 34. Auto-detect and optimize limited repetitions in JIT.
309 35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
310 blocking (*UTF) etc.
312 36. In the interpreter, maximizing pattern repetitions for characters and
313 character types now use tail recursion, which reduces stack usage.
315 37. The value of the max lookbehind was not correctly preserved if a compiled
316 and saved regex was reloaded on a host of different endianness.
318 38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
319 of the compiled pattern block, expand the flags field from 16 to 32 bits
320 because it was almost full.
322 39. Try madvise first before posix_madvise.
324 40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
325 with a pattern such as ^$. It has taken 4 years for anybody to notice! The
326 original change locked out all matches of empty strings. This has been
327 changed so that one match of an empty string per line is recognized.
328 Subsequent searches on the same line (for colouring or for --only-matching,
329 for example) do not recognize empty strings.
331 41. Applied a user patch to fix a number of spelling mistakes in comments.
333 42. Data lines longer than 65536 caused pcretest to crash.
335 43. Clarified the data type for length and startoffset arguments for pcre_exec
336 and pcre_dfa_exec in the function-specific man pages, where they were
337 explicitly stated to be in bytes, never having been updated. I also added
338 some clarification to the pcreapi man page.
340 44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
341 a segmentation fault.
344 Version 8.32 30-November-2012
345 -----------------------------
347 1. Improved JIT compiler optimizations for first character search and single
348 character iterators.
350 2. Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
351 Patch by Daniel Richard G.
353 3. Single character iterator optimizations in the JIT compiler.
355 4. Improved JIT compiler optimizations for character ranges.
357 5. Rename the "leave" variable names to "quit" to improve WinCE compatibility.
358 Reported by Giuseppe D'Angelo.
360 6. The PCRE_STARTLINE bit, indicating that a match can occur only at the start
361 of a line, was being set incorrectly in cases where .* appeared inside
362 atomic brackets at the start of a pattern, or where there was a subsequent
363 *PRUNE or *SKIP.
365 7. Improved instruction cache flush for POWER/PowerPC.
366 Patch by Daniel Richard G.
368 8. Fixed a number of issues in pcregrep, making it more compatible with GNU
369 grep:
371 (a) There is now no limit to the number of patterns to be matched.
373 (b) An error is given if a pattern is too long.
375 (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
376 are now supported.
378 (d) --exclude-from and --include-from (multiple use) have been added.
380 (e) Exclusions and inclusions now apply to all files and directories, not
381 just to those obtained from scanning a directory recursively.
383 (f) Multiple uses of -f and --file-list are now supported.
385 (g) In a Windows environment, the default for -d has been changed from
386 "read" (the GNU grep default) to "skip", because otherwise the presence
387 of a directory in the file list provokes an error.
389 (h) The documentation has been revised and clarified in places.
391 9. Improve the matching speed of capturing brackets.
393 10. Changed the meaning of \X so that it now matches a Unicode extended
394 grapheme cluster.
396 11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
397 out POSIX threads when JIT support is configured.
399 12. Added support for PCRE_STUDY_EXTRA_NEEDED.
401 13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
402 structure could go wrong in environments where size_t is not the same size
403 as int.
405 14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
406 set.
408 15. The EBCDIC support had decayed; later updates to the code had included
409 explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
410 general tidy up of EBCDIC-related issues, and the documentation was also
411 not quite right. There is now a test that can be run on ASCII systems to
412 check some of the EBCDIC-related things (but is it not a full test).
414 16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
415 in a small tidy to the code.
417 17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
419 18. If the --only-matching (-o) option in pcregrep is specified multiple
420 times, each one causes appropriate output. For example, -o1 -o2 outputs the
421 substrings matched by the 1st and 2nd capturing parentheses. A separating
422 string can be specified by --om-separator (default empty).
424 19. Improving the first n character searches.
426 20. Turn case lists for horizontal and vertical white space into macros so that
427 they are defined only once.
429 21. This set of changes together give more compatible Unicode case-folding
430 behaviour for characters that have more than one other case when UCP
431 support is available.
433 (a) The Unicode property table now has offsets into a new table of sets of
434 three or more characters that are case-equivalent. The MultiStage2.py
435 script that generates these tables (the pcre_ucd.c file) now scans
436 CaseFolding.txt instead of UnicodeData.txt for character case
437 information.
439 (b) The code for adding characters or ranges of characters to a character
440 class has been abstracted into a generalized function that also handles
441 case-independence. In UTF-mode with UCP support, this uses the new data
442 to handle characters with more than one other case.
444 (c) A bug that is fixed as a result of (b) is that codepoints less than 256
445 whose other case is greater than 256 are now correctly matched
446 caselessly. Previously, the high codepoint matched the low one, but not
447 vice versa.
449 (d) The processing of \h, \H, \v, and \ in character classes now makes use
450 of the new class addition function, using character lists defined as
451 macros alongside the case definitions of 20 above.
453 (e) Caseless back references now work with characters that have more than
454 one other case.
456 (f) General caseless matching of characters with more than one other case
457 is supported.
459 22. Unicode character properties were updated from Unicode 6.2.0
461 23. Improved CMake support under Windows. Patch by Daniel Richard G.
463 24. Add support for 32-bit character strings, and UTF-32
465 25. Major JIT compiler update (code refactoring and bugfixing).
466 Experimental Sparc 32 support is added.
468 26. Applied a modified version of Daniel Richard G's patch to create
469 pcre.h.generic and config.h.generic by "make" instead of in the
470 PrepareRelease script.
472 27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
473 pcre_compile.c when checking for a zero character.
475 28. Introducing a native interface for JIT. Through this interface, the compiled
476 machine code can be directly executed. The purpose of this interface is to
477 provide fast pattern matching, so several sanity checks are not performed.
478 However, feature tests are still performed. The new interface provides
479 1.4x speedup compared to the old one.
481 29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
482 the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
483 was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
485 30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
486 greater than 0x7fffffff (which cannot be represented in UTF-8, even under
487 the "old" RFC 2279). Instead, it ended up passing a negative length to
488 pcre_exec().
490 31. Add support for GCC's visibility feature to hide internal functions.
492 32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
493 "unknown -C option" after outputting 0 or 1.
495 33. There is now support for generating a code coverage report for the test
496 suite in environments where gcc is the compiler and lcov is installed. This
497 is mainly for the benefit of the developers.
499 34. If PCRE is built with --enable-valgrind, certain memory regions are marked
500 unaddressable using valgrind annotations, allowing valgrind to detect
501 invalid memory accesses. This is mainly for the benefit of the developers.
503 25. (*UTF) can now be used to start a pattern in any of the three libraries.
505 26. Give configure error if --enable-cpp but no C++ compiler found.
508 Version 8.31 06-July-2012
509 -------------------------
511 1. Fixing a wrong JIT test case and some compiler warnings.
513 2. Removed a bashism from the RunTest script.
515 3. Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
516 to unsigned type, result still unsigned" that was given by an MS compiler
517 on encountering the code "-sizeof(xxx)".
519 4. Partial matching support is added to the JIT compiler.
521 5. Fixed several bugs concerned with partial matching of items that consist
522 of more than one character:
524 (a) /^(..)\1/ did not partially match "aba" because checking references was
525 done on an "all or nothing" basis. This also applied to repeated
526 references.
528 (b) \R did not give a hard partial match if \r was found at the end of the
529 subject.
531 (c) \X did not give a hard partial match after matching one or more
532 characters at the end of the subject.
534 (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
535 a partial match for the string "\r".
537 (e) When newline was set to CRLF, the metacharacter "." did not recognize
538 a partial match for a CR character at the end of the subject string.
540 6. If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
541 running pcretest, the text "(JIT)" added to the output whenever JIT is
542 actually used to run the match.
544 7. Individual JIT compile options can be set in pcretest by following -s+[+]
545 or /S+[+] with a digit between 1 and 7.
547 8. OP_NOT now supports any UTF character not just single-byte ones.
549 9. (*MARK) control verb is now supported by the JIT compiler.
551 10. The command "./RunTest list" lists the available tests without actually
552 running any of them. (Because I keep forgetting what they all are.)
556 12. Applied a (slightly modified) user-supplied patch that improves performance
557 when the heap is used for recursion (compiled with --disable-stack-for-
558 recursion). Instead of malloc and free for each heap frame each time a
559 logical recursion happens, frames are retained on a chain and re-used where
560 possible. This sometimes gives as much as 30% improvement.
562 13. As documented, (*COMMIT) is now confined to within a recursive subpattern
563 call.
565 14. As documented, (*COMMIT) is now confined to within a positive assertion.
567 15. It is now possible to link pcretest with libedit as an alternative to
568 libreadline.
570 16. (*COMMIT) control verb is now supported by the JIT compiler.
572 17. The Unicode data tables have been updated to Unicode 6.1.0.
574 18. Added --file-list option to pcregrep.
576 19. Added binary file support to pcregrep, including the -a, --binary-files,
577 -I, and --text options.
579 20. The madvise function is renamed for posix_madvise for QNX compatibility
580 reasons. Fixed by Giuseppe D'Angelo.
582 21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
584 22. Changed the option for creating symbolic links for 16-bit man pages from
585 -s to -sf so that re-installing does not cause issues.
587 23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
589 24. Fixed a very old bug in pcretest that caused errors with restarted DFA
590 matches in certain environments (the workspace was not being correctly
591 retained). Also added to pcre_dfa_exec() a simple plausibility check on
592 some of the workspace data at the beginning of a restart.
594 25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
595 was not doing so when it should - probably a typo introduced by SVN 528
596 (change 8.10/14).
598 26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
599 \w+ when the character tables indicated that \x{c4} was a word character.
600 There were several related cases, all because the tests for doing a table
601 lookup were testing for characters less than 127 instead of 255.
603 27. If a pattern contains capturing parentheses that are not used in a match,
604 their slots in the ovector are set to -1. For those that are higher than
605 any matched groups, this happens at the end of processing. In the case when
606 there were back references that the ovector was too small to contain
607 (causing temporary malloc'd memory to be used during matching), and the
608 highest capturing number was not used, memory off the end of the ovector
609 was incorrectly being set to -1. (It was using the size of the temporary
610 memory instead of the true size.)
612 28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
613 ovector size, it uses memory at the end of the block that it has got.
615 29. Check for an overlong MARK name and give an error at compile time. The
616 limit is 255 for the 8-bit library and 65535 for the 16-bit library.
618 30. JIT compiler update.
620 31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
621 Rill for the patch.
623 32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
625 33. Variable renamings in the PCRE-JIT compiler. No functionality change.
627 34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
628 SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
629 was enabled.
631 35. Improve JIT code generation for greedy plus quantifier.
633 36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
634 1 to "aa" instead of to an empty string. The bug affected repeated groups
635 that could potentially match an empty string.
637 37. Optimizing single character iterators in JIT.
639 38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
640 the same checks as \x{...} characters in non-JavaScript mode. Specifically,
641 codepoints that are too big for the mode are faulted, and in a UTF mode,
642 disallowed codepoints are also faulted.
644 39. If PCRE was compiled with UTF support, in three places in the DFA
645 matcher there was code that should only have been obeyed in UTF mode, but
646 was being obeyed unconditionally. In 8-bit mode this could cause incorrect
647 processing when bytes with values greater than 127 were present. In 16-bit
648 mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
649 both cases the values are those that cannot be the first data item in a UTF
650 character. The three items that might have provoked this were recursions,
651 possessively repeated groups, and atomic groups.
653 40. Ensure that libpcre is explicitly listed in the link commands for pcretest
654 and pcregrep, because some OS require shared objects to be explicitly
655 passed to ld, causing the link step to fail if they are not.
657 41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
658 mode, patterns that started with \h* or \R* might be incorrectly matched.
661 Version 8.30 04-February-2012
662 -----------------------------
664 1. Renamed "isnumber" as "is_a_number" because in some Mac environments this
665 name is defined in ctype.h.
667 2. Fixed a bug in fixed-length calculation for lookbehinds that would show up
668 only in quite long subpatterns.
670 3. Removed the function pcre_info(), which has been obsolete and deprecated
671 since it was replaced by pcre_fullinfo() in February 2000.
673 4. For a non-anchored pattern, if (*SKIP) was given with a name that did not
674 match a (*MARK), and the match failed at the start of the subject, a
675 reference to memory before the start of the subject could occur. This bug
676 was introduced by fix 17 of release 8.21.
678 5. A reference to an unset group with zero minimum repetition was giving
679 totally wrong answers (in non-JavaScript-compatibility mode). For example,
680 /(another)?(\1?)test/ matched against "hello world test". This bug was
681 introduced in release 8.13.
683 6. Add support for 16-bit character strings (a large amount of work involving
684 many changes and refactorings).
686 7. RunGrepTest failed on msys because \r\n was replaced by whitespace when the
687 command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
688 from a file.
690 8. Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
691 rounding is not applied in this particular case).
693 9. The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
694 if they appear, or are escaped, in patterns.
696 10. Get rid of a number of -Wunused-but-set-variable warnings.
698 11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
699 "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
700 Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
701 also returns the mark "x". This bug applied to capturing parentheses,
702 non-capturing parentheses, and atomic parentheses. It also applied to some
703 assertions.
705 12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
706 information out of configure.ac instead of relying on pcre.h.generic, which
707 is not stored in the repository.
709 13. Applied Dmitry V. Levin's patch for a more portable method for linking with
710 -lreadline.
712 14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
714 15. Applied Graycode's patch to put the top-level frame on the stack rather
715 than the heap when not using the stack for recursion. This gives a
716 performance improvement in many cases when recursion is not deep.
718 16. Experimental code added to "pcretest -C" to output the stack frame size.
721 Version 8.21 12-Dec-2011
722 ------------------------
724 1. Updating the JIT compiler.
726 2. JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
727 are added as well.
729 3. Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
730 PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
731 calling _pcre_jit_exec. Some extra comments are added.
733 4. (*MARK) settings inside atomic groups that do not contain any capturing
734 parentheses, for example, (?>a(*:m)), were not being passed out. This bug
735 was introduced by change 18 for 8.20.
737 5. Supporting of \x, \U and \u in JavaScript compatibility mode based on the
738 ECMA-262 standard.
740 6. Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
741 erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
742 This bug was probably introduced by change 9 of 8.13.
744 7. While fixing 6 above, I noticed that a number of other items were being
745 incorrectly rejected as "not fixed length". This arose partly because newer
746 opcodes had not been added to the fixed-length checking code. I have (a)
747 corrected the bug and added tests for these items, and (b) arranged for an
748 error to occur if an unknown opcode is encountered while checking for fixed
749 length instead of just assuming "not fixed length". The items that were
750 rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
751 (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
752 repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
754 8. A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
755 being incorrectly compiled and would have given unpredicatble results.
757 9. A possessively repeated subpattern with minimum repeat count greater than
758 one behaved incorrectly. For example, (A){2,}+ behaved as if it was
759 (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
760 the first (A) could occur when it should not.
762 10. Add a cast and remove a redundant test from the code.
764 11. JIT should use pcre_malloc/pcre_free for allocation.
766 12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
767 best practice nowadays, and helps with cross-compiling. (If the exec_prefix
768 is anything other than /usr, -L is still shown).
770 13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
772 14. Perl does not support \N without a following name in a [] class; PCRE now
773 also gives an error.
775 15. If a forward reference was repeated with an upper limit of around 2000,
776 it caused the error "internal error: overran compiling workspace". The
777 maximum number of forward references (including repeats) was limited by the
778 internal workspace, and dependent on the LINK_SIZE. The code has been
779 rewritten so that the workspace expands (via pcre_malloc) if necessary, and
780 the default depends on LINK_SIZE. There is a new upper limit (for safety)
781 of around 200,000 forward references. While doing this, I also speeded up
782 the filling in of repeated forward references.
784 16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
785 incorrectly expecting the subject to contain another "a" after the start.
787 17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
788 in the match, the SKIP should be ignored. This was not happening; instead
789 the SKIP was being treated as NOMATCH. For patterns such as
790 /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
791 tested.
793 18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
794 now much more compatible with Perl, in particular in cases where the result
795 is a non-match for a non-anchored pattern. For example, if
796 /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
797 "m", where previously it did not return a name. A side effect of this
798 change is that for partial matches, the last encountered mark name is
799 returned, as for non matches. A number of tests that were previously not
800 Perl-compatible have been moved into the Perl-compatible test files. The
801 refactoring has had the pleasing side effect of removing one argument from
802 the match() function, thus reducing its stack requirements.
804 19. If the /S+ option was used in pcretest to study a pattern using JIT,
805 subsequent uses of /S (without +) incorrectly behaved like /S+.
807 21. Retrieve executable code size support for the JIT compiler and fixing
808 some warnings.
810 22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
811 not work when the shorter character appeared right at the end of the
812 subject string.
814 23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
815 systems.
817 24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
818 output it when the /M option is used in pcretest.
820 25. The CheckMan script was not being included in the distribution. Also, added
821 an explicit "perl" to run Perl scripts from the PrepareRelease script
822 because this is reportedly needed in Windows.
824 26. If study data was being save in a file and studying had not found a set of
825 "starts with" bytes for the pattern, the data written to the file (though
826 never used) was taken from uninitialized memory and so caused valgrind to
827 complain.
829 27. Updated RunTest.bat as provided by Sheri Pierce.
831 28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
833 29. Computation of memory usage for the table of capturing group names was
834 giving an unnecessarily large value.
837 Version 8.20 21-Oct-2011
838 ------------------------
840 1. Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
841 a POSIX class. After further experiments with Perl, which convinced me that
842 Perl has bugs and confusions, a closing square bracket is no longer allowed
843 in a POSIX name. This bug also affected patterns with classes that started
844 with full stops.
846 2. If a pattern such as /(a)b|ac/ is matched against "ac", there is no
847 captured substring, but while checking the failing first alternative,
848 substring 1 is temporarily captured. If the output vector supplied to
849 pcre_exec() was not big enough for this capture, the yield of the function
850 was still zero ("insufficient space for captured substrings"). This cannot
851 be totally fixed without adding another stack variable, which seems a lot
852 of expense for a edge case. However, I have improved the situation in cases
853 such as /(a)(b)x|abc/ matched against "abc", where the return code
854 indicates that fewer than the maximum number of slots in the ovector have
855 been set.
857 3. Related to (2) above: when there are more back references in a pattern than
858 slots in the output vector, pcre_exec() uses temporary memory during
859 matching, and copies in the captures as far as possible afterwards. It was
860 using the entire output vector, but this conflicts with the specification
861 that only 2/3 is used for passing back captured substrings. Now it uses
862 only the first 2/3, for compatibility. This is, of course, another edge
863 case.
865 4. Zoltan Herczeg's just-in-time compiler support has been integrated into the
866 main code base, and can be used by building with --enable-jit. When this is
867 done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
868 runtime --no-jit option is given.
870 5. When the number of matches in a pcre_dfa_exec() run exactly filled the
871 ovector, the return from the function was zero, implying that there were
872 other matches that did not fit. The correct "exactly full" value is now
873 returned.
875 6. If a subpattern that was called recursively or as a subroutine contained
876 (*PRUNE) or any other control that caused it to give a non-standard return,
877 invalid errors such as "Error -26 (nested recursion at the same subject
878 position)" or even infinite loops could occur.
880 7. If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
881 computing the minimum length on reaching *ACCEPT, and so ended up with the
882 wrong value of 1 rather than 0. Further investigation indicates that
883 computing a minimum subject length in the presence of *ACCEPT is difficult
884 (think back references, subroutine calls), and so I have changed the code
885 so that no minimum is registered for a pattern that contains *ACCEPT.
887 8. If (*THEN) was present in the first (true) branch of a conditional group,
888 it was not handled as intended. [But see 16 below.]
890 9. Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
891 Sheri Pierce.
893 10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
894 the first byte in a match must be "a".
896 11. Change 17 for 8.13 increased the recursion depth for patterns like
897 /a(?:.)*?a/ drastically. I've improved things by remembering whether a
898 pattern contains any instances of (*THEN). If it does not, the old
899 optimizations are restored. It would be nice to do this on a per-group
900 basis, but at the moment that is not feasible.
902 12. In some environments, the output of pcretest -C is CRLF terminated. This
903 broke RunTest's code that checks for the link size. A single white space
904 character after the value is now allowed for.
906 13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
907 For "fr", it uses the Windows-specific input and output files.
909 14. If (*THEN) appeared in a group that was called recursively or as a
910 subroutine, it did not work as intended. [But see next item.]
912 15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
913 pattern fragments (but not containing any | characters). If A and B are
914 matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
915 was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
916 D. In other words, Perl considers parentheses that do not contain any |
917 characters to be part of a surrounding alternative, whereas PCRE was
918 treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
919 differently. PCRE now behaves in the same way as Perl, except in the case
920 of subroutine/recursion calls such as (?1) which have in any case always
921 been different (but PCRE had them first :-).
923 16. Related to 15 above: Perl does not treat the | in a conditional group as
924 creating alternatives. Such a group is treated in the same way as an
925 ordinary group without any | characters when processing (*THEN). PCRE has
926 been changed to match Perl's behaviour.
928 17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
929 RunGrepTest script failed.
931 18. Change 22 for version 13 caused atomic groups to use more stack. This is
932 inevitable for groups that contain captures, but it can lead to a lot of
933 stack use in large patterns. The old behaviour has been restored for atomic
934 groups that do not contain any capturing parentheses.
936 19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
937 suppress the check for a minimum subject length at run time. (If it was
938 given to pcre_exec() or pcre_dfa_exec() it did work.)
940 20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
941 fail to work when decoding hex characters in data strings in EBCDIC
942 environments.
944 21. It appears that in at least one Mac OS environment, the isxdigit() function
945 is implemented as a macro that evaluates to its argument more than once,
946 contravening the C 90 Standard (I haven't checked a later standard). There
947 was an instance in pcretest which caused it to go wrong when processing
948 \x{...} escapes in subject strings. The has been rewritten to avoid using
949 things like p++ in the argument of isxdigit().
952 Version 8.13 16-Aug-2011
953 ------------------------
955 1. The Unicode data tables have been updated to Unicode 6.0.0.
957 2. Two minor typos in pcre_internal.h have been fixed.
959 3. Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
960 pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
961 in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
963 4. There were a number of related bugs in the code for matching backrefences
964 caselessly in UTF-8 mode when codes for the characters concerned were
965 different numbers of bytes. For example, U+023A and U+2C65 are an upper
966 and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
967 (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
968 code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
969 2-byte code at the end of the subject (it thought there wasn't enough data
970 left).
972 5. Comprehensive information about what went wrong is now returned by
973 pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
974 as the output vector has at least 2 elements. The offset of the start of
975 the failing character and a reason code are placed in the vector.
977 6. When the UTF-8 string check fails for pcre_compile(), the offset that is
978 now returned is for the first byte of the failing character, instead of the
979 last byte inspected. This is an incompatible change, but I hope it is small
980 enough not to be a problem. It makes the returned offset consistent with
981 pcre_exec() and pcre_dfa_exec().
983 7. pcretest now gives a text phrase as well as the error number when
984 pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
985 failure, the offset and reason code are output.
987 8. When \R was used with a maximizing quantifier it failed to skip backwards
988 over a \r\n pair if the subsequent match failed. Instead, it just skipped
989 back over a single character (\n). This seems wrong (because it treated the
990 two characters as a single entity when going forwards), conflicts with the
991 documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
992 behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
993 has been changed.
995 9. Some internal refactoring has changed the processing so that the handling
996 of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
997 time (the PCRE_DOTALL option was changed this way some time ago: version
998 7.7 change 16). This has made it possible to abolish the OP_OPT op code,
999 which was always a bit of a fudge. It also means that there is one less
1000 argument for the match() function, which reduces its stack requirements
1001 slightly. This change also fixes an incompatibility with Perl: the pattern
1002 (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1004 10. More internal refactoring has drastically reduced the number of recursive
1005 calls to match() for possessively repeated groups such as (abc)++ when
1006 using pcre_exec().
1008 11. While implementing 10, a number of bugs in the handling of groups were
1009 discovered and fixed:
1011 (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1012 (a|)*(?1) gave a compile-time internal error.
1013 ((a|)+)+ did not notice that the outer group could match an empty string.
1014 (^a|^)+ was not marked as anchored.
1015 (.*a|.*)+ was not marked as matching at start or after a newline.
1017 12. Yet more internal refactoring has removed another argument from the match()
1018 function. Special calls to this function are now indicated by setting a
1019 value in a variable in the "match data" data block.
1021 13. Be more explicit in pcre_study() instead of relying on "default" for
1022 opcodes that mean there is no starting character; this means that when new
1023 ones are added and accidentally left out of pcre_study(), testing should
1024 pick them up.
1026 14. The -s option of pcretest has been documented for ages as being an old
1027 synonym of -m (show memory usage). I have changed it to mean "force study
1028 for every regex", that is, assume /S for every regex. This is similar to -i
1029 and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1030 using it. It makes it easier to run collections of tests with and without
1031 study enabled, and thereby test pcre_study() more easily. All the standard
1032 tests are now run with and without -s (but some patterns can be marked as
1033 "never study" - see 20 below).
1035 15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1036 restoration of the capturing data to the outer values was not happening
1037 correctly.
1039 16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1040 empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1041 pattern had matched an empty string, and so incorrectly returned a no
1042 match.
1044 17. There was optimizing code for the last branch of non-capturing parentheses,
1045 and also for the obeyed branch of a conditional subexpression, which used
1046 tail recursion to cut down on stack usage. Unfortunately, now that there is
1047 the possibility of (*THEN) occurring in these branches, tail recursion is
1048 no longer possible because the return has to be checked for (*THEN). These
1049 two optimizations have therefore been removed. [But see 8.20/11 above.]
1051 18. If a pattern containing \R was studied, it was assumed that \R always
1052 matched two bytes, thus causing the minimum subject length to be
1053 incorrectly computed because \R can also match just one byte.
1055 19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1056 was incorrectly computed.
1058 20. If /S is present twice on a test pattern in pcretest input, it now
1059 *disables* studying, thereby overriding the use of -s on the command line
1060 (see 14 above). This is necessary for one or two tests to keep the output
1061 identical in both cases.
1063 21. When (*ACCEPT) was used in an assertion that matched an empty string and
1064 PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1066 22. When an atomic group that contained a capturing parenthesis was
1067 successfully matched, but the branch in which it appeared failed, the
1068 capturing was not being forgotten if a higher numbered group was later
1069 captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1070 group 1 to "a", when in fact it should be unset. This applied to multi-
1071 branched capturing and non-capturing groups, repeated or not, and also to
1072 positive assertions (capturing in negative assertions does not happen
1073 in PCRE) and also to nested atomic groups.
1075 23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1076 subject after a captured substring, to make it easier to tell which of a
1077 number of identical substrings has been captured.
1079 24. The way atomic groups are processed by pcre_exec() has been changed so that
1080 if they are repeated, backtracking one repetition now resets captured
1081 values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1082 "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1083 "aaa". Previously, it would have been "a". As part of this code
1084 refactoring, the way recursive calls are handled has also been changed.
1086 25. If an assertion condition captured any substrings, they were not passed
1087 back unless some other capturing happened later. For example, if
1088 (?(?=(a))a) was matched against "a", no capturing was returned.
1090 26. When studying a pattern that contained subroutine calls or assertions,
1091 the code for finding the minimum length of a possible match was handling
1092 direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1093 group 1 called group 2 while simultaneously a separate group 2 called group
1094 1). A stack overflow occurred in this case. I have fixed this by limiting
1095 the recursion depth to 10.
1097 27. Updated RunTest.bat in the distribution to the version supplied by Tom
1098 Fortmann. This supports explicit test numbers on the command line, and has
1099 argument validation and error reporting.
1101 28. An instance of \X with an unlimited repeat could fail if at any point the
1102 first character it looked at was a mark character.
1104 29. Some minor code refactoring concerning Unicode properties and scripts
1105 should reduce the stack requirement of match() slightly.
1107 30. Added the '=' option to pcretest to check the setting of unused capturing
1108 slots at the end of the pattern, which are documented as being -1, but are
1109 not included in the return count.
1111 31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1112 compiled something random. Now it gives a compile-time error (as does
1113 Perl).
1115 32. A *MARK encountered during the processing of a positive assertion is now
1116 recorded and passed back (compatible with Perl).
1118 33. If --only-matching or --colour was set on a pcregrep call whose pattern
1119 had alternative anchored branches, the search for a second match in a line
1120 was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1121 matched the line "0102" twice. The same bug affected patterns that started
1122 with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1123 twice.
1125 34. Previously, PCRE did not allow quantification of assertions. However, Perl
1126 does, and because of capturing effects, quantifying parenthesized
1127 assertions may at times be useful. Quantifiers are now allowed for
1128 parenthesized assertions.
1130 35. A minor code tidy in pcre_compile() when checking options for \R usage.
1132 36. \g was being checked for fancy things in a character class, when it should
1133 just be a literal "g".
1135 37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1136 appearance of a nested POSIX class supersedes an apparent external class.
1137 For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1138 unescaped square brackets may also appear as part of class names. For
1139 example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1140 more like Perl. (But see 8.20/1 above.)
1142 38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1143 was because it thought it was \N{name}, which is not supported).
1145 39. Add minix to OS list not supporting the -S option in pcretest.
1147 40. PCRE tries to detect cases of infinite recursion at compile time, but it
1148 cannot analyze patterns in sufficient detail to catch mutual recursions
1149 such as ((?1))((?2)). There is now a runtime test that gives an error if a
1150 subgroup is called recursively as a subpattern for a second time at the
1151 same position in the subject string. In previous releases this might have
1152 been caught by the recursion limit, or it might have run out of stack.
1154 41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1155 happen only once. PCRE was, however incorrectly giving a compile time error
1156 "recursive call could loop indefinitely" because it cannot analyze the
1157 pattern in sufficient detail. The compile time test no longer happens when
1158 PCRE is compiling a conditional subpattern, but actual runaway loops are
1159 now caught at runtime (see 40 above).
1161 42. It seems that Perl allows any characters other than a closing parenthesis
1162 to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1163 has been changed to be the same.
1165 43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1166 as not to get warnings when autogen.sh is called. Also changed
1167 AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1169 44. To help people who use pcregrep to scan files containing exceedingly long
1170 lines, the following changes have been made:
1172 (a) The default value of the buffer size parameter has been increased from
1173 8K to 20K. (The actual buffer used is three times this size.)
1175 (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1176 PCRE is built.
1178 (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1179 to be set at run time.
1181 (d) Numerical values in pcregrep options can be followed by K or M, for
1182 example --buffer-size=50K.
1184 (e) If a line being scanned overflows pcregrep's buffer, an error is now
1185 given and the return code is set to 2.
1187 45. Add a pointer to the latest mark to the callout data block.
1189 46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1190 partial match of an empty string instead of no match. This was specific to
1191 the use of ".".
1193 47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1194 complete match instead of a partial match. This bug was dependent on both
1195 the PCRE_UTF8 and PCRE_DOTALL options being set.
1197 48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1198 starting byte set, because \b was not being ignored.
1201 Version 8.12 15-Jan-2011
1202 ------------------------
1204 1. Fixed some typos in the markup of the man pages, and wrote a script that
1205 checks for such things as part of the documentation building process.
1207 2. On a big-endian 64-bit system, pcregrep did not correctly process the
1208 --match-limit and --recursion-limit options (added for 8.11). In
1209 particular, this made one of the standard tests fail. (The integer value
1210 went into the wrong half of a long int.)
1212 3. If the --colour option was given to pcregrep with -v (invert match), it
1213 did strange things, either producing crazy output, or crashing. It should,
1214 of course, ignore a request for colour when reporting lines that do not
1215 match.
1217 4. Another pcregrep bug caused similar problems if --colour was specified with
1218 -M (multiline) and the pattern match finished with a line ending.
1220 5. In pcregrep, when a pattern that ended with a literal newline sequence was
1221 matched in multiline mode, the following line was shown as part of the
1222 match. This seems wrong, so I have changed it.
1224 6. Another pcregrep bug in multiline mode, when --colour was specified, caused
1225 the check for further matches in the same line (so they could be coloured)
1226 to overrun the end of the current line. If another match was found, it was
1227 incorrectly shown (and then shown again when found in the next line).
1229 7. If pcregrep was compiled under Windows, there was a reference to the
1230 function pcregrep_exit() before it was defined. I am assuming this was
1231 the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1232 reported by a user. I've moved the definition above the reference.
1235 Version 8.11 10-Dec-2010
1236 ------------------------
1238 1. (*THEN) was not working properly if there were untried alternatives prior
1239 to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1240 backtracked to try for "b" instead of moving to the next alternative branch
1241 at the same level (in this case, to look for "c"). The Perl documentation
1242 is clear that when (*THEN) is backtracked onto, it goes to the "next
1243 alternative in the innermost enclosing group".
1245 2. (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1246 such as (A(*COMMIT)B(*THEN)C|D) any failure after matching A should
1247 result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1248 (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1249 (*THEN).
1251 3. If \s appeared in a character class, it removed the VT character from
1252 the class, even if it had been included by some previous item, for example
1253 in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1254 of \s, but is part of the POSIX "space" class.)
1256 4. A partial match never returns an empty string (because you can always
1257 match an empty string at the end of the subject); however the checking for
1258 an empty string was starting at the "start of match" point. This has been
1259 changed to the "earliest inspected character" point, because the returned
1260 data for a partial match starts at this character. This means that, for
1261 example, /(?<=abc)def/ gives a partial match for the subject "abc"
1262 (previously it gave "no match").
1264 5. Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1265 of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1266 previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1267 has an implication that the given string is incomplete (because a partial
1268 match is preferred over a full match). For this reason, these items now
1269 give a partial match in this situation. [Aside: previously, the one case
1270 /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1271 match rather than a full match, which was wrong by the old rules, but is
1272 now correct.]
1274 6. There was a bug in the handling of #-introduced comments, recognized when
1275 PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1276 If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1277 UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1278 scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1279 but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1280 places in pcre_compile().
1282 7. Related to (6) above, when pcre_compile() was skipping #-introduced
1283 comments when looking ahead for named forward references to subpatterns,
1284 the only newline sequence it recognized was NL. It now handles newlines
1285 according to the set newline convention.
1287 8. SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1288 former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1289 cater for a lack of strerror(). These oversights have been fixed.
1291 9. Added --match-limit and --recursion-limit to pcregrep.
1293 10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1295 11. When the -o option was used, pcregrep was setting a return code of 1, even
1296 when matches were found, and --line-buffered was not being honoured.
1298 12. Added an optional parentheses number to the -o and --only-matching options
1299 of pcregrep.
1301 13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1302 can match an empty string. The code to do it in pcretest and pcredemo
1303 needed fixing:
1305 (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1306 only one byte after an empty string match just before CRLF (this case
1307 just got forgotten; "any" and "anycrlf" were OK).
1309 (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1310 mode when an empty string match preceded an ASCII character followed by
1311 a non-ASCII character. (The code for advancing by one character rather
1312 than one byte was nonsense.)
1314 (c) The pcredemo.c sample program did not have any code at all to handle
1315 the cases when CRLF is a valid newline sequence.
1317 14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1318 as a starting offset was within the subject string. There is now a new
1319 error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1320 negative or greater than the length of the string. In order to test this,
1321 pcretest is extended to allow the setting of negative starting offsets.
1323 15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1324 starting offset points to the beginning of a UTF-8 character was
1325 unnecessarily clumsy. I tidied it up.
1327 16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1328 bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1330 17. Nobody had reported that the --include_dir option, which was added in
1331 release 7.7 should have been called --include-dir (hyphen, not underscore)
1332 for compatibility with GNU grep. I have changed it to --include-dir, but
1333 left --include_dir as an undocumented synonym, and the same for
1334 --exclude-dir, though that is not available in GNU grep, at least as of
1335 release 2.5.4.
1337 18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1338 characters from a string of bytes) have been redefined so as not to use
1339 loops, in order to improve performance in some environments. At the same
1340 time, I abstracted some of the common code into auxiliary macros to save
1341 repetition (this should not affect the compiled code).
1343 19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1344 compile-time error is now given if \c is not followed by an ASCII
1345 character, that is, a byte less than 128. (In EBCDIC mode, the code is
1346 different, and any byte value is allowed.)
1348 20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1349 START_OPTIMIZE option, which is now allowed at compile time - but just
1350 passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1351 to pcregrep and other applications that have no direct access to PCRE
1352 options. The new /Y option in pcretest sets this option when calling
1353 pcre_compile().
1355 21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1356 back references. Groups containing recursive back references were forced to
1357 be atomic by that change, but in the case of named groups, the amount of
1358 memory required was incorrectly computed, leading to "Failed: internal
1359 error: code overflow". This has been fixed.
1361 22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1362 pcretest.c, to avoid build problems in some Borland environments.
1365 Version 8.10 25-Jun-2010
1366 ------------------------
1368 1. Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1369 THEN.
1371 2. (*ACCEPT) was not working when inside an atomic group.
1373 3. Inside a character class, \B is treated as a literal by default, but
1374 faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1375 causes the error). The code is unchanged, but I tidied the documentation.
1377 4. Inside a character class, PCRE always treated \R and \X as literals,
1378 whereas Perl faults them if its -w option is set. I have changed PCRE so
1379 that it faults them when PCRE_EXTRA is set.
1381 5. Added support for \N, which always matches any character other than
1382 newline. (It is the same as "." when PCRE_DOTALL is not set.)
1384 6. When compiling pcregrep with newer versions of gcc which may have
1385 FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1386 declared with attribute warn_unused_result" were given. Just casting the
1387 result to (void) does not stop the warnings; a more elaborate fudge is
1388 needed. I've used a macro to implement this.
1390 7. Minor change to pcretest.c to avoid a compiler warning.
1392 8. Added four artifical Unicode properties to help with an option to make
1393 \s etc use properties (see next item). The new properties are: Xan
1394 (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1396 9. Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1397 use Unicode properties. (*UCP) at the start of a pattern can be used to set
1398 this option. Modified pcretest to add /W to test this facility. Added
1399 REG_UCP to make it available via the POSIX interface.
1401 10. Added --line-buffered to pcregrep.
1403 11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1404 studied, and the match started with a letter with a code point greater than
1405 127 whose first byte was different to the first byte of the other case of
1406 the letter, the other case of this starting letter was not recognized
1407 (#976).
1409 12. If a pattern that was studied started with a repeated Unicode property
1410 test, for example, \p{Nd}+, there was the theoretical possibility of
1411 setting up an incorrect bitmap of starting bytes, but fortunately it could
1412 not have actually happened in practice until change 8 above was made (it
1413 added property types that matched character-matching opcodes).
1415 13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1416 possible starting bytes for non-anchored patterns.
1418 14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1419 \R, and also a number of cases that involve Unicode properties, both
1420 explicit and implicit when PCRE_UCP is set.
1422 15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1423 input, it could crash or give wrong results if characters with values
1424 greater than 0xc0 were present in the subject string. (Detail: it assumed
1425 UTF-8 input when processing these items.)
1427 16. Added a lot of (int) casts to avoid compiler warnings in systems where
1428 size_t is 64-bit (#991).
1430 17. Added a check for running out of memory when PCRE is compiled with
1431 --disable-stack-for-recursion (#990).
1433 18. If the last data line in a file for pcretest does not have a newline on
1434 the end, a newline was missing in the output.
1436 19. The default pcre_chartables.c file recognizes only ASCII characters (values
1437 less than 128) in its various bitmaps. However, there is a facility for
1438 generating tables according to the current locale when PCRE is compiled. It
1439 turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1440 characters, are recognized by isspace() and therefore were getting set in
1441 these tables, and indeed these tables seem to approximate to ISO 8859. This
1442 caused a problem in UTF-8 mode when pcre_study() was used to create a list
1443 of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1444 which of course cannot start UTF-8 characters. I have changed the code so
1445 that only real ASCII characters (less than 128) and the correct starting
1446 bytes for UTF-8 encodings are set for characters greater than 127 when in
1447 UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1448 altogether.)
1450 20. Added the /T option to pcretest so as to be able to run tests with non-
1451 standard character tables, thus making it possible to include the tests
1452 used for 19 above in the standard set of tests.
1454 21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1455 reference to a subpattern the other side of a comment that contains an
1456 opening parenthesis caused either an internal compiling error, or a
1457 reference to the wrong subpattern.
1460 Version 8.02 19-Mar-2010
1461 ------------------------
1463 1. The Unicode data tables have been updated to Unicode 5.2.0.
1465 2. Added the option --libs-cpp to pcre-config, but only when C++ support is
1466 configured.
1468 3. Updated the licensing terms in the pcregexp.pas file, as agreed with the
1469 original author of that file, following a query about its status.
1471 4. On systems that do not have stdint.h (e.g. Solaris), check for and include
1472 inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1474 5. A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1475 quantifier applied to a forward-referencing subroutine call, could compile
1476 incorrect code or give the error "internal error: previously-checked
1477 referenced subpattern not found".
1479 6. Both MS Visual Studio and Symbian OS have problems with initializing
1480 variables to point to external functions. For these systems, therefore,
1481 pcre_malloc etc. are now initialized to local functions that call the
1482 relevant global functions.
1484 7. There were two entries missing in the vectors called coptable and poptable
1485 in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1486 I've fixed the data, and added a kludgy way of testing at compile time that
1487 the lengths are correct (equal to the number of opcodes).
1489 8. Following on from 7, I added a similar kludge to check the length of the
1490 eint vector in pcreposix.c.
1492 9. Error texts for pcre_compile() are held as one long string to avoid too
1493 much relocation at load time. To find a text, the string is searched,
1494 counting zeros. There was no check for running off the end of the string,
1495 which could happen if a new error number was added without updating the
1496 string.
1498 10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1500 11. \K was not working if it appeared in an atomic group or in a group that
1501 was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1502 \K is "not well defined" if used in an assertion. PCRE now accepts it if
1503 the assertion is positive, but not if it is negative.
1505 12. Change 11 fortuitously reduced the size of the stack frame used in the
1506 "match()" function of pcre_exec.c by one pointer. Forthcoming
1507 implementation of support for (*MARK) will need an extra pointer on the
1508 stack; I have reserved it now, so that the stack frame size does not
1509 decrease.
1511 13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1512 item in branch that calls a recursion is a subroutine call - as in the
1513 second branch in the above example - was incorrectly given the compile-
1514 time error "recursive call could loop indefinitely" because pcre_compile()
1515 was not correctly checking the subroutine for matching a non-empty string.
1517 14. The checks for overrunning compiling workspace could trigger after an
1518 overrun had occurred. This is a "should never occur" error, but it can be
1519 triggered by pathological patterns such as hundreds of nested parentheses.
1520 The checks now trigger 100 bytes before the end of the workspace.
1522 15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1525 Version 8.01 19-Jan-2010
1526 ------------------------
1528 1. If a pattern contained a conditional subpattern with only one branch (in
1529 particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1530 computed the wrong minimum data length (which is of course zero for such
1531 subpatterns). This could cause incorrect "no match" results.
1533 2. For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1534 the pattern is reset in the first branch, pcre_compile() failed with
1535 "internal error: code overflow at offset...". This happened only when
1536 the reset was to the original external option setting. (An optimization
1537 abstracts leading options settings into an external setting, which was the
1538 cause of this.)
1540 3. A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1541 of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1542 assertion pattern did not match (meaning that the assertion was true), it
1543 was incorrectly treated as false if the SKIP had been reached during the
1544 matching. This also applied to assertions used as conditions.
1546 4. If an item that is not supported by pcre_dfa_exec() was encountered in an
1547 assertion subpattern, including such a pattern used as a condition,
1548 unpredictable results occurred, instead of the error return
1551 5. The C++ GlobalReplace function was not working like Perl for the special
1552 situation when an empty string is matched. It now does the fancy magic
1553 stuff that is necessary.
1555 6. In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1556 removed. (These were left over from very, very early versions of PCRE.)
1558 7. Some cosmetic changes to the code to make life easier when compiling it
1559 as part of something else:
1561 (a) Change DEBUG to PCRE_DEBUG.
1563 (b) In pcre_compile(), rename the member of the "branch_chain" structure
1564 called "current" as "current_branch", to prevent a collision with the
1565 Linux macro when compiled as a kernel module.
1567 (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1568 prevent a collision with the Linux macro when compiled as a kernel
1569 module.
1571 8. In pcre_compile() there are some checks for integer overflows that used to
1572 cast potentially large values to (double). This has been changed to that
1573 when building, a check for int64_t is made, and if it is found, it is used
1574 instead, thus avoiding the use of floating point arithmetic. (There is no
1575 other use of FP in PCRE.) If int64_t is not found, the fallback is to
1576 double.
1578 9. Added two casts to avoid signed/unsigned warnings from VS Studio Express
1579 2005 (difference between two addresses compared to an unsigned value).
1581 10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1582 custom one, because of the following reported problem in Windows:
1584 - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1585 under Win32.
1586 - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1587 therefore missing the function definition.
1588 - The compiler thus generates a "C" signature for the test function.
1589 - The linker fails to find the "C" function.
1590 - PCRE fails to configure if asked to do so against libbz2.
1592 11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1593 messages were output:
1595 Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1596 rerunning libtoolize, to keep the correct libtool macros in-tree.
1597 Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1599 I have done both of these things.
1601 12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1602 most of the time, it *can* run out if it is given a pattern that contains a
1603 runaway infinite recursion. I updated the discussion in the pcrestack man
1604 page.
1606 13. Now that we have gone to the x.xx style of version numbers, the minor
1607 version may start with zero. Using 08 or 09 is a bad idea because users
1608 might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1609 interpreted as invalid octal numbers. I've updated the previous comment in
1610 configure.ac, and also added a check that gives an error if 08 or 09 are
1611 used.
1613 14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1614 causing partial matching to fail when the end of the subject matched \W
1615 in a UTF-8 pattern where \W was quantified with a minimum of 3.
1617 15. There were some discrepancies between the declarations in pcre_internal.h
1618 of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1619 their definitions. The declarations used "const uschar *" and the
1620 definitions used USPTR. Even though USPTR is normally defined as "const
1621 unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1622 reported that: "This difference in casting confuses some C++ compilers, for
1623 example, SunCC recognizes above declarations as different functions and
1624 generates broken code for hbpcre." I have changed the declarations to use
1625 USPTR.
1627 16. GNU libtool is named differently on some systems. The autogen.sh script now
1628 tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1629 (FreeBSD).
1631 17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1632 (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1633 comment: "Figure out how to create a longlong from a string: strtoll and
1634 equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1635 instance, but it only takes 2 args instead of 3!"
1637 18. A subtle bug concerned with back references has been fixed by a change of
1638 specification, with a corresponding code fix. A pattern such as
1639 ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1640 refers, was giving matches when it shouldn't. For example, xa=xaaa would
1641 match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1642 same bug. Such groups have to be quantified to be useful, or contained
1643 inside another quantified group. (If there's no repetition, the reference
1644 can never match.) The problem arises because, having left the group and
1645 moved on to the rest of the pattern, a later failure that backtracks into
1646 the group uses the captured value from the final iteration of the group
1647 rather than the correct earlier one. I have fixed this in PCRE by forcing
1648 any group that contains a reference to itself to be an atomic group; that
1649 is, there cannot be any backtracking into it once it has completed. This is
1650 similar to recursive and subroutine calls.
1653 Version 8.00 19-Oct-09
1654 ----------------------
1656 1. The table for translating pcre_compile() error codes into POSIX error codes
1657 was out-of-date, and there was no check on the pcre_compile() error code
1658 being within the table. This could lead to an OK return being given in
1659 error.
1661 2. Changed the call to open a subject file in pcregrep from fopen(pathname,
1662 "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1663 in a Windows environment.
1665 3. The pcregrep --count option prints the count for each file even when it is
1666 zero, as does GNU grep. However, pcregrep was also printing all files when
1667 --files-with-matches was added. Now, when both options are given, it prints
1668 counts only for those files that have at least one match. (GNU grep just
1669 prints the file name in this circumstance, but including the count seems
1670 more useful - otherwise, why use --count?) Also ensured that the
1671 combination -clh just lists non-zero counts, with no names.
1673 4. The long form of the pcregrep -F option was incorrectly implemented as
1674 --fixed_strings instead of --fixed-strings. This is an incompatible change,
1675 but it seems right to fix it, and I didn't think it was worth preserving
1676 the old behaviour.
1678 5. The command line items --regex=pattern and --regexp=pattern were not
1679 recognized by pcregrep, which required --regex pattern or --regexp pattern
1680 (with a space rather than an '='). The man page documented the '=' forms,
1681 which are compatible with GNU grep; these now work.
1683 6. No libpcreposix.pc file was created for pkg-config; there was just
1684 libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1686 7. Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1687 when UCP support is not needed, by modifying the Python script that
1688 generates it from Unicode data files. This should not matter if the module
1689 is correctly used as a library, but I received one complaint about 50K of
1690 unwanted data. My guess is that the person linked everything into his
1691 program rather than using a library. Anyway, it does no harm.
1693 8. A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1694 was a minimum greater than 1 for a wide character in a possessive
1695 repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1696 which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1697 character. Chaos in the form of incorrect output or a compiling loop could
1698 result.
1700 9. The restrictions on what a pattern can contain when partial matching is
1701 requested for pcre_exec() have been removed. All patterns can now be
1702 partially matched by this function. In addition, if there are at least two
1703 slots in the offset vector, the offset of the earliest inspected character
1704 for the match and the offset of the end of the subject are set in them when
1705 PCRE_ERROR_PARTIAL is returned.
1707 10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1708 synonymous with PCRE_PARTIAL, for backwards compatibility, and
1709 PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1710 and may be more useful for multi-segment matching.
1712 11. Partial matching with pcre_exec() is now more intuitive. A partial match
1713 used to be given if ever the end of the subject was reached; now it is
1714 given only if matching could not proceed because another character was
1715 needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1716 string "Z", which now yields "no match" instead of "partial match". In the
1717 case of pcre_dfa_exec(), "no match" is given if every matching path for the
1718 final character ended with (*FAIL).
1720 12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1721 if the pattern had a "must contain" character that was already found in the
1722 earlier partial match, unless partial matching was again requested. For
1723 example, with the pattern /dog.(body)?/, the "must contain" character is
1724 "g". If the first part-match was for the string "dog", restarting with
1725 "sbody" failed. This bug has been fixed.
1727 13. The string returned by pcre_dfa_exec() after a partial match has been
1728 changed so that it starts at the first inspected character rather than the
1729 first character of the match. This makes a difference only if the pattern
1730 starts with a lookbehind assertion or \b or \B (\K is not supported by
1731 pcre_dfa_exec()). It's an incompatible change, but it makes the two
1732 matching functions compatible, and I think it's the right thing to do.
1734 14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1735 so that the demonstration program is easily available in environments where
1736 PCRE has not been installed from source.
1738 15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1739 libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1740 library.
1742 16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1743 It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1744 is not the first non-POSIX option to be added. Clearly some people find
1745 these options useful.
1747 17. If a caller to the POSIX matching function regexec() passes a non-zero
1748 value for nmatch with a NULL value for pmatch, the value of
1749 nmatch is forced to zero.
1751 18. RunGrepTest did not have a test for the availability of the -u option of
1752 the diff command, as RunTest does. It now checks in the same way as
1753 RunTest, and also checks for the -b option.
1755 19. If an odd number of negated classes containing just a single character
1756 interposed, within parentheses, between a forward reference to a named
1757 subpattern and the definition of the subpattern, compilation crashed with
1758 an internal error, complaining that it could not find the referenced
1759 subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1760 [The bug was that it was starting one character too far in when skipping
1761 over the character class, thus treating the ] as data rather than
1762 terminating the class. This meant it could skip too much.]
1764 20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1765 /g option in pcretest when the pattern contains \K, which makes it possible
1766 to have an empty string match not at the start, even when the pattern is
1767 anchored. Updated pcretest and pcredemo to use this option.
1769 21. If the maximum number of capturing subpatterns in a recursion was greater
1770 than the maximum at the outer level, the higher number was returned, but
1771 with unset values at the outer level. The correct (outer level) value is
1772 now given.
1774 22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1775 PCRE did not set those parentheses (unlike Perl). I have now found a way to
1776 make it do so. The string so far is captured, making this feature
1777 compatible with Perl.
1779 23. The tests have been re-organized, adding tests 11 and 12, to make it
1780 possible to check the Perl 5.10 features against Perl 5.10.
1782 24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1783 pattern matches a fixed length string. PCRE did not allow this; now it
1784 does. Neither allows recursion.
1786 25. I finally figured out how to implement a request to provide the minimum
1787 length of subject string that was needed in order to match a given pattern.
1788 (It was back references and recursion that I had previously got hung up
1789 on.) This code has now been added to pcre_study(); it finds a lower bound
1790 to the length of subject needed. It is not necessarily the greatest lower
1791 bound, but using it to avoid searching strings that are too short does give
1792 some useful speed-ups. The value is available to calling programs via
1793 pcre_fullinfo().
1795 26. While implementing 25, I discovered to my embarrassment that pcretest had
1796 not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1797 study optimizations had never been tested with that matching function.
1798 Oops. What is worse, even when it was passed study data, there was a bug in
1799 pcre_dfa_exec() that meant it never actually used it. Double oops. There
1800 were also very few tests of studied patterns with pcre_dfa_exec().
1802 27. If (?| is used to create subpatterns with duplicate numbers, they are now
1803 allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1804 on the other side of the coin, they are no longer allowed to have different
1805 names, because these cannot be distinguished in PCRE, and this has caused
1806 confusion. (This is a difference from Perl.)
1808 28. When duplicate subpattern names are present (necessarily with different
1809 numbers, as required by 27 above), and a test is made by name in a
1810 conditional pattern, either for a subpattern having been matched, or for
1811 recursion in such a pattern, all the associated numbered subpatterns are
1812 tested, and the overall condition is true if the condition is true for any
1813 one of them. This is the way Perl works, and is also more like the way
1814 testing by number works.
1817 Version 7.9 11-Apr-09
1818 ---------------------
1820 1. When building with support for bzlib/zlib (pcregrep) and/or readline
1821 (pcretest), all targets were linked against these libraries. This included
1822 libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1823 libraries. This caused unwanted dependencies to be created. This problem
1824 has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1825 pcretest is linked with readline.
1827 2. The "typedef int BOOL" in pcre_internal.h that was included inside the
1828 "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1829 moved outside it again, because FALSE and TRUE are already defined in AIX,
1830 but BOOL is not.
1832 3. The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1833 PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1835 4. The pcregrep documentation said spaces were inserted as well as colons (or
1836 hyphens) following file names and line numbers when outputting matching
1837 lines. This is not true; no spaces are inserted. I have also clarified the
1838 wording for the --colour (or --color) option.
1840 5. In pcregrep, when --colour was used with -o, the list of matching strings
1841 was not coloured; this is different to GNU grep, so I have changed it to be
1842 the same.
1844 6. When --colo(u)r was used in pcregrep, only the first matching substring in
1845 each matching line was coloured. Now it goes on to look for further matches
1846 of any of the test patterns, which is the same behaviour as GNU grep.
1848 7. A pattern that could match an empty string could cause pcregrep to loop; it
1849 doesn't make sense to accept an empty string match in pcregrep, so I have
1850 locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1851 seems to be how GNU grep behaves. [But see later change 40 for release
1852 8.33.]
1854 8. The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1855 start or after a newline", because the conditional assertion was not being
1856 correctly handled. The rule now is that both the assertion and what follows
1857 in the first alternative must satisfy the test.
1859 9. If auto-callout was enabled in a pattern with a conditional group whose
1860 condition was an assertion, PCRE could crash during matching, both with
1861 pcre_exec() and pcre_dfa_exec().
1863 10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1864 used for matching.
1866 11. Unicode property support in character classes was not working for
1867 characters (bytes) greater than 127 when not in UTF-8 mode.
1869 12. Added the -M command line option to pcretest.
1871 14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1873 15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1875 16. Added comments and documentation about mis-use of no_arg in the C++
1876 wrapper.
1878 17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1879 from Martin Jerabek that uses macro names for all relevant character and
1880 string constants.
1882 18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1883 SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1884 SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1885 these, but not everybody uses configure.
1887 19. A conditional group that had only one branch was not being correctly
1888 recognized as an item that could match an empty string. This meant that an
1889 enclosing group might also not be so recognized, causing infinite looping
1890 (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1891 with the subject "ab", where knowledge that the repeated group can match
1892 nothing is needed in order to break the loop.
1894 20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1895 exec(), but without supplying a callout function, matching went wrong.
1897 21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1898 leak if the size of the offset vector was greater than 30. When the vector
1899 is smaller, the saved offsets during recursion go onto a local stack
1900 vector, but for larger vectors malloc() is used. It was failing to free
1901 when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1902 error, in fact).
1904 22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1905 heapframe that is used only when UTF-8 support is enabled. This caused no
1906 problem, but was untidy.
1908 23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1909 CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1910 included within another project.
1912 24. Steven Van Ingelgem's patches to add more options to the CMake support,
1913 slightly modified by me:
1915 (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1916 not building pcregrep.
1918 (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1919 if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1921 25. Forward references, both numeric and by name, in patterns that made use of
1922 duplicate group numbers, could behave incorrectly or give incorrect errors,
1923 because when scanning forward to find the reference group, PCRE was not
1924 taking into account the duplicate group numbers. A pattern such as
1925 ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1927 26. Changed a few more instances of "const unsigned char *" to USPTR, making
1928 the feature of a custom pointer more persuasive (as requested by a user).
1930 27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1931 pcretest.c, inside #ifndefs, because it seems they are sometimes already
1932 pre-defined.
1934 28. Added support for (*UTF8) at the start of a pattern.
1936 29. Arrange for flags added by the "release type" setting in CMake to be shown
1937 in the configuration summary.
1940 Version 7.8 05-Sep-08
1941 ---------------------
1943 1. Replaced UCP searching code with optimized version as implemented for Ad
1944 Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1945 stage table and inline lookup instead of a function, giving speed ups of 2
1946 to 5 times on some simple patterns that I tested. Permission was given to
1947 distribute the MultiStage2.py script that generates the tables (it's not in
1948 the tarball, but is in the Subversion repository).
1950 2. Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1951 scripts.
1953 3. Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1954 a group with a zero qualifier. The result of the study could be incorrect,
1955 or the function might crash, depending on the pattern.
1957 4. Caseless matching was not working for non-ASCII characters in back
1958 references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1959 It now works when Unicode Property Support is available.
1961 5. In pcretest, an escape such as \x{de} in the data was always generating
1962 a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1963 non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1964 truncation.
1966 6. Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1968 7. Added two (int) casts to pcregrep when printing the difference of two
1969 pointers, in case they are 64-bit values.
1971 8. Added comments about Mac OS X stack usage to the pcrestack man page and to
1972 test 2 if it fails.
1974 9. Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1975 and a #define of that name to empty if it is not externally set. This is to
1976 allow users of MSVC to set it if necessary.
1978 10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1979 the convenience functions in the pcre_get.c source file.
1981 11. An option change at the start of a pattern that had top-level alternatives
1982 could cause overwriting and/or a crash. This command provoked a crash in
1983 some environments:
1985 printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1987 This potential security problem was recorded as CVE-2008-2371.
1989 12. For a pattern where the match had to start at the beginning or immediately
1990 after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
1991 pcre_dfa_exec() could read past the end of the passed subject if there was
1992 no match. To help with detecting such bugs (e.g. with valgrind), I modified
1993 pcretest so that it places the subject at the end of its malloc-ed buffer.
1995 13. The change to pcretest in 12 above threw up a couple more cases when pcre_
1996 exec() might read past the end of the data buffer in UTF-8 mode.
1998 14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
1999 the data contained the byte 0x85 as part of a UTF-8 character within its
2000 first line. This applied both to normal and DFA matching.
2002 15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2003 /^[^d]*?$/8 failed to match "abc".
2005 16. Added a missing copyright notice to pcrecpp_internal.h.
2007 17. Make it more clear in the documentation that values returned from
2008 pcre_exec() in ovector are byte offsets, not character counts.
2010 18. Tidied a few places to stop certain compilers from issuing warnings.
2012 19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2013 supplied by Stefan Weber. I made a further small update for 7.8 because
2014 there is a change of source arrangements: the pcre_searchfuncs.c module is
2015 replaced by pcre_ucd.c.
2018 Version 7.7 07-May-08
2019 ---------------------
2021 1. Applied Craig's patch to sort out a long long problem: "If we can't convert
2022 a string to a long long, pretend we don't even have a long long." This is
2023 done by checking for the strtoq, strtoll, and _strtoi64 functions.
2025 2. Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2026 pre-7.6 versions, which defined a global no_arg variable instead of putting
2027 it in the RE class. (See also #8 below.)
2029 3. Remove a line of dead code, identified by coverity and reported by Nuno
2030 Lopes.
2032 4. Fixed two related pcregrep bugs involving -r with --include or --exclude:
2034 (1) The include/exclude patterns were being applied to the whole pathnames
2035 of files, instead of just to the final components.
2037 (2) If there was more than one level of directory, the subdirectories were
2038 skipped unless they satisfied the include/exclude conditions. This is
2039 inconsistent with GNU grep (and could even be seen as contrary to the
2040 pcregrep specification - which I improved to make it absolutely clear).
2041 The action now is always to scan all levels of directory, and just
2042 apply the include/exclude patterns to regular files.
2044 5. Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2045 --exclude_dir in the tests to avoid scanning .svn directories.
2047 6. Applied Craig's patch to the QuoteMeta function so that it escapes the
2048 NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2049 doesn't support NULs in patterns.
2051 7. Added some missing "const"s to declarations of static tables in
2052 pcre_compile.c and pcre_dfa_exec.c.
2054 8. Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2055 caused by fix #2 above. (Subsequently also a second patch to fix the
2056 first patch. And a third patch - this was a messy problem.)
2058 9. Applied Craig's patch to remove the use of push_back().
2060 10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2061 matching function regexec().
2063 11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2064 which, however, unlike Perl's \g{...}, are subroutine calls, not back
2065 references. PCRE supports relative numbers with this syntax (I don't think
2066 Oniguruma does).
2068 12. Previously, a group with a zero repeat such as (...){0} was completely
2069 omitted from the compiled regex. However, this means that if the group
2070 was called as a subroutine from elsewhere in the pattern, things went wrong
2071 (an internal error was given). Such groups are now left in the compiled
2072 pattern, with a new opcode that causes them to be skipped at execution
2073 time.
2075 13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2076 to the way PCRE behaves:
2078 (a) A lone ] character is dis-allowed (Perl treats it as data).
2080 (b) A back reference to an unmatched subpattern matches an empty string
2081 (Perl fails the current match path).
2083 (c) A data ] in a character class must be notated as \] because if the
2084 first data character in a class is ], it defines an empty class. (In
2085 Perl it is not possible to have an empty class.) The empty class []
2086 never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2087 The negative empty class [^] matches any one character, independently
2088 of the DOTALL setting.
2090 14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2091 non-existent subpattern following a character class starting with ']' and
2092 containing () gave an internal compiling error instead of "reference to
2093 non-existent subpattern". Fortunately, when the pattern did exist, the
2094 compiled code was correct. (When scanning forwards to check for the
2095 existence of the subpattern, it was treating the data ']' as terminating
2096 the class, so got the count wrong. When actually compiling, the reference
2097 was subsequently set up correctly.)
2099 15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2100 it was being rejected as not supported by pcre_dfa_exec(), even though
2101 other assertions are supported. I have made pcre_dfa_exec() support
2102 (*FAIL).
2104 16. The implementation of 13c above involved the invention of a new opcode,
2105 OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2106 cannot be changed at match time, I realized I could make a small
2107 improvement to matching performance by compiling OP_ALLANY instead of
2108 OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2109 on the OP_ANY path.
2111 17. Compiling pcretest on Windows with readline support failed without the
2112 following two fixes: (1) Make the unistd.h include conditional on
2113 HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2115 18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2116 ncurses library to be included for pcretest when ReadLine support is
2117 requested, but also to allow for it to be overridden. This patch came from
2118 Daniel Bergström.
2120 19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2121 as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2122 any errors with the current Unicode tables. Thanks to Peter Kankowski for
2123 spotting this.
2126 Version 7.6 28-Jan-08
2127 ---------------------
2129 1. A character class containing a very large number of characters with
2130 codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2131 overflow.
2133 2. Patch to cut out the "long long" test in pcrecpp_unittest when
2134 HAVE_LONG_LONG is not defined.
2136 3. Applied Christian Ehrlicher's patch to update the CMake build files to
2137 bring them up to date and include new features. This patch includes:
2139 - Fixed PH's badly added libz and libbz2 support.
2140 - Fixed a problem with static linking.
2141 - Added pcredemo. [But later removed - see 7 below.]
2142 - Fixed dftables problem and added an option.
2143 - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2145 - Added readline support for pcretest.
2146 - Added an listing of the option settings after cmake has run.
2148 4. A user submitted a patch to Makefile that makes it easy to create
2149 "pcre.dll" under mingw when using Configure/Make. I added stuff to
2150 Makefile.am that cause it to include this special target, without
2151 affecting anything else. Note that the same mingw target plus all
2152 the other distribution libraries and programs are now supported
2153 when configuring with CMake (see 6 below) instead of with
2154 Configure/Make.
2156 5. Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2157 This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2158 exported in the Windows port". It has not yet been confirmed that the patch
2159 solves the problem, but it does no harm.
2161 6. Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2162 NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2163 with CMake, and also correct the comment about stack recursion.
2165 7. Remove the automatic building of pcredemo from the ./configure system and
2166 from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2167 of a program that users should build themselves after PCRE is installed, so
2168 building it automatically is not really right. What is more, it gave
2169 trouble in some build environments.
2171 8. Further tidies to CMakeLists.txt from Sheri and Christian.
2174 Version 7.5 10-Jan-08
2175 ---------------------
2177 1. Applied a patch from Craig: "This patch makes it possible to 'ignore'
2178 values in parens when parsing an RE using the C++ wrapper."
2180 2. Negative specials like \S did not work in character classes in UTF-8 mode.
2181 Characters greater than 255 were excluded from the class instead of being
2182 included.
2184 3. The same bug as (2) above applied to negated POSIX classes such as
2185 [:^space:].
2187 4. PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2188 defined or documented. It seems to have been a typo for PCRE_STATIC, so
2189 I have changed it.
2191 5. The construct (?&) was not diagnosed as a syntax error (it referenced the
2192 first named subpattern) and a construct such as (?&a) would reference the
2193 first named subpattern whose name started with "a" (in other words, the
2194 length check was missing). Both these problems are fixed. "Subpattern name
2195 expected" is now given for (?&) (a zero-length name), and this patch also
2196 makes it give the same error for \k'' (previously it complained that that
2197 was a reference to a non-existent subpattern).
2199 6. The erroneous patterns (?+-a) and (?-+a) give different error messages;
2200 this is right because (?- can be followed by option settings as well as by
2201 digits. I have, however, made the messages clearer.
2203 7. Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2204 than the number used in the conditional) now cause a compile-time error.
2205 This is actually not compatible with Perl, which accepts such patterns, but
2206 treats the conditional as always being FALSE (as PCRE used to), but it
2207 seems to me that giving a diagnostic is better.
2209 8. Change "alphameric" to the more common word "alphanumeric" in comments
2210 and messages.
2212 9. Fix two occurrences of "backslash" in comments that should have been
2213 "backspace".
2215 10. Remove two redundant lines of code that can never be obeyed (their function
2216 was moved elsewhere).
2218 11. The program that makes PCRE's Unicode character property table had a bug
2219 which caused it to generate incorrect table entries for sequences of
2220 characters that have the same character type, but are in different scripts.
2221 It amalgamated them into a single range, with the script of the first of
2222 them. In other words, some characters were in the wrong script. There were
2223 thirteen such cases, affecting characters in the following ranges:
2225 U+002b0 - U+002c1
2226 U+0060c - U+0060d
2227 U+0061e - U+00612
2228 U+0064b - U+0065e
2229 U+0074d - U+0076d
2230 U+01800 - U+01805
2231 U+01d00 - U+01d77
2232 U+01d9b - U+01dbf
2233 U+0200b - U+0200f
2234 U+030fc - U+030fe
2235 U+03260 - U+0327f
2236 U+0fb46 - U+0fbb1
2237 U+10450 - U+1049d
2239 12. The -o option (show only the matching part of a line) for pcregrep was not
2240 compatible with GNU grep in that, if there was more than one match in a
2241 line, it showed only the first of them. It now behaves in the same way as
2242 GNU grep.
2244 13. If the -o and -v options were combined for pcregrep, it printed a blank
2245 line for every non-matching line. GNU grep prints nothing, and pcregrep now
2246 does the same. The return code can be used to tell if there were any
2247 non-matching lines.
2249 14. Added --file-offsets and --line-offsets to pcregrep.
2251 15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2252 infinitely looping recursion. The bug was that positive lookaheads were not
2253 being skipped when checking for a possible empty match (negative lookaheads
2254 and both kinds of lookbehind were skipped).
2256 16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2257 inclusion of <windows.h> to before rather than after the definition of
2258 INVALID_FILE_ATTRIBUTES (patch from David Byron).
2260 17. Specifying a possessive quantifier with a specific limit for a Unicode
2261 character property caused pcre_compile() to compile bad code, which led at
2262 runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2263 are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2264 caused the error; without that there was no problem.
2266 18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2268 19. Added --enable-pcretest-libreadline.
2270 20. In pcrecpp.cc, the variable 'count' was incremented twice in
2271 RE::GlobalReplace(). As a result, the number of replacements returned was
2272 double what it should be. I removed one of the increments, but Craig sent a
2273 later patch that removed the other one (the right fix) and added unit tests
2274 that check the return values (which was not done before).
2276 21. Several CMake things:
2278 (1) Arranged that, when cmake is used on Unix, the libraries end up with
2279 the names libpcre and libpcreposix, not just pcre and pcreposix.
2281 (2) The above change means that pcretest and pcregrep are now correctly
2282 linked with the newly-built libraries, not previously installed ones.
2286 22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2287 crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2288 UTF-8 newline character). The key issue is that the pattern starts .*;
2289 this means that the match must be either at the beginning, or after a
2290 newline. The bug was in the code for advancing after a failed match and
2291 checking that the new position followed a newline. It was not taking
2292 account of UTF-8 characters correctly.
2294 23. PCRE was behaving differently from Perl in the way it recognized POSIX
2295 character classes. PCRE was not treating the sequence [:...:] as a
2296 character class unless the ... were all letters. Perl, however, seems to
2297 allow any characters between [: and :], though of course it rejects as
2298 unknown any "names" that contain non-letters, because all the known class
2299 names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2300 for example, whereas PCRE did not - it did not recognize a POSIX character
2301 class. This seemed a bit dangerous, so the code has been changed to be
2302 closer to Perl. The behaviour is not identical to Perl, because PCRE will
2303 diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2304 treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2305 Perl does, and where it didn't before.
2307 24. Rewrite so as to remove the single use of %n from pcregrep because in some
2308 Windows environments %n is disabled by default.
2311 Version 7.4 21-Sep-07
2312 ---------------------
2314 1. Change 7.3/28 was implemented for classes by looking at the bitmap. This
2315 means that a class such as [\s] counted as "explicit reference to CR or
2316 LF". That isn't really right - the whole point of the change was to try to
2317 help when there was an actual mention of one of the two characters. So now
2318 the change happens only if \r or \n (or a literal CR or LF) character is
2319 encountered.
2321 2. The 32-bit options word was also used for 6 internal flags, but the numbers
2322 of both had grown to the point where there were only 3 bits left.
2323 Fortunately, there was spare space in the data structure, and so I have
2324 moved the internal flags into a new 16-bit field to free up more option
2325 bits.
2327 3. The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2328 but did not set the internal JCHANGED flag - either of these is enough to
2329 control the way the "get" function works - but the PCRE_INFO_JCHANGED
2330 facility is supposed to tell if (?J) was ever used, so now (?J) at the
2331 start sets both bits.
2333 4. Added options (at build time, compile time, exec time) to change \R from
2334 matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2336 5. doc/pcresyntax.html was missing from the distribution.
2338 6. Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2339 compatibility, even though it is no longer used.
2341 7. Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2342 strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2343 windows.h file is present (where different names are used). [This was
2344 reversed later after testing - see 16 below.]
2346 8. Changed all #include <config.h> to #include "config.h". There were also
2347 some further <pcre.h> cases that I changed to "pcre.h".
2349 9. When pcregrep was used with the --colour option, it missed the line ending
2350 sequence off the lines that it output.
2352 10. It was pointed out to me that arrays of string pointers cause lots of
2353 relocations when a shared library is dynamically loaded. A technique of
2354 using a single long string with a table of offsets can drastically reduce
2355 these. I have refactored PCRE in four places to do this. The result is
2356 dramatic:
2358 Originally: 290
2359 After changing UCP table: 187
2360 After changing error message table: 43
2361 After changing table of "verbs" 36
2362 After changing table of Posix names 22
2364 Thanks to the folks working on Gregex for glib for this insight.
2366 11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2367 unicode-properties was also set.
2369 12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2371 13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2372 checked only for CRLF.
2374 14. Added casts to pcretest.c to avoid compiler warnings.
2376 15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2378 16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2379 and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2380 entirely. This removes changes made in 7 above.
2382 17. The CMake files have been updated, and there is now more information about
2383 building with CMake in the NON-UNIX-USE document.
2386 Version 7.3 28-Aug-07
2387 ---------------------
2389 1. In the rejigging of the build system that eventually resulted in 7.1, the
2390 line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2391 brackets there is not right, since it causes compilers to look for an
2392 installed pcre.h, not the version that is in the source that is being
2393 compiled (which of course may be different). I have changed it back to:
2395 #include "pcre.h"
2397 I have a vague recollection that the change was concerned with compiling in
2398 different directories, but in the new build system, that is taken care of
2399 by the VPATH setting the Makefile.
2401 2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2402 when the subject happened to end in the byte 0x85 (e.g. if the last
2403 character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2404 characters but of course it shouldn't be taken as a newline when it is part
2405 of another character. The bug was that, for an unlimited repeat of . in
2406 not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2407 characters when looking for a newline.
2409 3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2411 4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2412 in debug output.
2414 5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2415 long printing in the pcrecpp unittest when running under MinGW.
2417 6. ESC_K was left out of the EBCDIC table.
2419 7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2420 parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2421 limit also applies to "virtual nesting" when a pattern is recursive, and in
2422 this case 1000 isn't so big. I have been able to remove this limit at the
2423 expense of backing off one optimization in certain circumstances. Normally,
2424 when pcre_exec() would call its internal match() function recursively and
2425 immediately return the result unconditionally, it uses a "tail recursion"
2426 feature to save stack. However, when a subpattern that can match an empty
2427 string has an unlimited repetition quantifier, it no longer makes this
2428 optimization. That gives it a stack frame in which to save the data for
2429 checking that an empty string has been matched. Previously this was taken
2430 from the 1000-entry workspace that had been reserved. So now there is no
2431 explicit limit, but more stack is used.
2433 8. Applied Daniel's patches to solve problems with the import/export magic
2434 syntax that is required for Windows, and which was going wrong for the
2435 pcreposix and pcrecpp parts of the library. These were overlooked when this
2436 problem was solved for the main library.
2438 9. There were some crude static tests to avoid integer overflow when computing
2439 the size of patterns that contain repeated groups with explicit upper
2440 limits. As the maximum quantifier is 65535, the maximum group length was
2441 set at 30,000 so that the product of these two numbers did not overflow a
2442 32-bit integer. However, it turns out that people want to use groups that
2443 are longer than 30,000 bytes (though not repeat them that many times).
2444 Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2445 made it possible to implement the integer overflow checks in a much more
2446 dynamic way, which I have now done. The artificial limitation on group
2447 length has been removed - we now have only the limit on the total length of
2448 the compiled pattern, which depends on the LINK_SIZE setting.
2450 10. Fixed a bug in the documentation for get/copy named substring when
2451 duplicate names are permitted. If none of the named substrings are set, the
2452 functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2453 empty string.
2455 11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2456 instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2457 because the ] is interpreted as the first data character and the
2458 terminating ] is not found. PCRE has been made compatible with Perl in this
2459 regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2460 cause memory overwriting.
2462 10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2463 string has been matched (to stop an infinite loop). It was not recognizing
2464 a conditional subpattern that could match an empty string if that
2465 subpattern was within another subpattern. For example, it looped when
2466 trying to match (((?(1)X|))*) but it was OK with ((?(1)X|)*) where the
2467 condition was not nested. This bug has been fixed.
2469 12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2470 past the start of the subject in the presence of bytes with the top bit
2471 set, for example "\x8aBCD".
2473 13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2474 (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2476 14. Optimized (?!) to (*FAIL).
2478 15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2479 This restricts code points to be within the range 0 to 0x10FFFF, excluding
2480 the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2481 full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2482 does: it's just the validity check that is more restrictive.
2484 16. Inserted checks for integer overflows during escape sequence (backslash)
2485 processing, and also fixed erroneous offset values for syntax errors during
2486 backslash processing.
2488 17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2489 for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2491 18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2492 caused an overrun.
2494 19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2495 something other than just ASCII characters) inside a group that had an
2496 unlimited repeat caused a loop at compile time (while checking to see
2497 whether the group could match an empty string).
2499 20. Debugging a pattern containing \p or \P could cause a crash. For example,
2500 [\P{Any}] did so. (Error in the code for printing property names.)
2502 21. An orphan \E inside a character class could cause a crash.
2504 22. A repeated capturing bracket such as (A)? could cause a wild memory
2505 reference during compilation.
2507 23. There are several functions in pcre_compile() that scan along a compiled
2508 expression for various reasons (e.g. to see if it's fixed length for look
2509 behind). There were bugs in these functions when a repeated \p or \P was
2510 present in the pattern. These operators have additional parameters compared
2511 with \d, etc, and these were not being taken into account when moving along
2512 the compiled data. Specifically:
2514 (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2515 length.
2517 (b) An item such as \pL+ within a repeated group could cause crashes or
2518 loops.
2520 (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2521 "reference to non-existent subpattern" error.
2523 (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2525 24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2526 characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2528 25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2530 26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2531 character were causing crashes (broken optimization).
2533 27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2534 \p or \P) caused a compile-time loop.
2536 28. More problems have arisen in unanchored patterns when CRLF is a valid line
2537 break. For example, the unstudied pattern [\r\n]A does not match the string
2538 "\r\nA" because change 7.0/46 below moves the current point on by two
2539 characters after failing to match at the start. However, the pattern \nA
2540 *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2541 the same is true. There doesn't seem any very clean way out of this, but
2542 what I have chosen to do makes the common cases work: PCRE now takes note
2543 of whether there can be an explicit match for \r or \n anywhere in the
2544 pattern, and if so, 7.0/46 no longer applies. As part of this change,
2545 there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2546 pattern has explicit CR or LF references.
2548 29. Added (*CR) etc for changing newline setting at start of pattern.
2551 Version 7.2 19-Jun-07
2552 ---------------------
2554 1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2555 which is apparently normally available under Windows.
2557 2. Re-jig the pcregrep tests with different newline settings in an attempt
2558 to make them independent of the local environment's newline setting.
2560 3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2562 4. Some of the "internals" tests were previously cut out when the link size
2563 was not 2, because the output contained actual offsets. The recent new
2564 "Z" feature of pcretest means that these can be cut out, making the tests
2565 usable with all link sizes.
2567 5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2568 stack recursion. This gives a massive performance boost under BSD, but just
2569 a small improvement under Linux. However, it saves one field in the frame
2570 in all cases.
2572 6. Added more features from the forthcoming Perl 5.10:
2574 (a) (?-n) (where n is a string of digits) is a relative subroutine or
2575 recursion call. It refers to the nth most recently opened parentheses.
2577 (b) (?+n) is also a relative subroutine call; it refers to the nth next
2578 to be opened parentheses.
2580 (c) Conditions that refer to capturing parentheses can be specified
2581 relatively, for example, (?(-2)... or (?(+3)...
2583 (d) \K resets the start of the current match so that everything before
2584 is not part of it.
2586 (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2588 (f) \g{name} is another synonym - part of Perl 5.10's unification of
2589 reference syntax.
2591 (g) (?| introduces a group in which the numbering of parentheses in each
2592 alternative starts with the same number.
2594 (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2596 7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2599 8. A pattern such as (.*(.)?)* caused pcre_exec() to fail by either not
2600 terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2601 for detecting groups that can match an empty string.
2603 9. A pattern with a very large number of alternatives (more than several
2604 hundred) was running out of internal workspace during the pre-compile
2605 phase, where pcre_compile() figures out how much memory will be needed. A
2606 bit of new cunning has reduced the workspace needed for groups with
2607 alternatives. The 1000-alternative test pattern now uses 12 bytes of
2608 workspace instead of running out of the 4096 that are available.
2610 10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2612 11. Applied patch from Google to remove an optimization that didn't quite work.
2613 The report of the bug said:
2615 pcrecpp::RE("a*").FullMatch("aaa") matches, while
2616 pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2617 pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2619 12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2620 it matched the wrong number of bytes.
2623 Version 7.1 24-Apr-07
2624 ---------------------
2626 1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2627 that is more "standard", making use of automake and other Autotools. There
2628 is some re-arrangement of the files and adjustment of comments consequent
2629 on this.
2631 2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2632 for recursive directory scanning broke on some systems because the files
2633 are not scanned in any specific order and on different systems the order
2634 was different. A call to "sort" has been inserted into RunGrepTest for the
2635 approprate test as a short-term fix. In the longer term there may be an
2636 alternative.
2638 3. I had an email from Eric Raymond about problems translating some of PCRE's
2639 man pages to HTML (despite the fact that I distribute HTML pages, some
2640 people do their own conversions for various reasons). The problems
2641 concerned the use of low-level troff macros .br and .in. I have therefore
2642 removed all such uses from the man pages (some were redundant, some could
2643 be replaced by .nf/.fi pairs). The 132html script that I use to generate
2644 HTML has been updated to handle .nf/.fi and to complain if it encounters
2645 .br or .in.
2647 4. Updated comments in configure.ac that get placed in config.h.in and also
2648 arranged for config.h to be included in the distribution, with the name
2649 config.h.generic, for the benefit of those who have to compile without
2650 Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2652 5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2653 Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2654 makevp.bat for the current PCRE, using the additional files
2655 makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2657 6. A Windows user reported a minor discrepancy with test 2, which turned out
2658 to be caused by a trailing space on an input line that had got lost in his
2659 copy. The trailing space was an accident, so I've just removed it.
2661 7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2662 that is needed.
2664 8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2665 as "const" (a) because they are and (b) because it helps the PHP
2666 maintainers who have recently made a script to detect big data structures
2667 in the php code that should be moved to the .rodata section. I remembered
2668 to update Builducptable as well, so it won't revert if ucptable.h is ever
2669 re-created.
2671 9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2672 pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2673 order to be able to cut out the UTF-8 tables in the latter when UTF-8
2674 support is not required. This saves 1.5-2K of code, which is important in
2675 some applications.
2677 Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2678 so as not to refer to the tables, even though these functions will never be
2679 called when UTF-8 support is disabled. Otherwise there are problems with a
2680 shared library.
2682 10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2684 (a) It was defining its arguments as char * instead of void *.
2686 (b) It was assuming that all moves were upwards in memory; this was true
2687 a long time ago when I wrote it, but is no longer the case.
2689 The emulated memove() is provided for those environments that have neither
2690 memmove() nor bcopy(). I didn't think anyone used it these days, but that
2691 is clearly not the case, as these two bugs were recently reported.
2693 11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2694 and Detrail to create the HTML documentation, the .txt form of the man
2695 pages, and it removes trailing spaces from listed files. It also creates
2696 pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2697 case, it wraps all the #defines with #ifndefs. This script should be run
2698 before "make dist".
2700 12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2701 with Unicode property support.
2703 (a) For a maximizing quantifier, if the two different cases of the
2704 character were of different lengths in their UTF-8 codings (there are
2705 some cases like this - I found 11), and the matching function had to
2706 back up over a mixture of the two cases, it incorrectly assumed they
2707 were both the same length.
2709 (b) When PCRE was configured to use the heap rather than the stack for
2710 recursion during matching, it was not correctly preserving the data for
2711 the other case of a UTF-8 character when checking ahead for a match
2712 while processing a minimizing repeat. If the check also involved
2713 matching a wide character, but failed, corruption could cause an
2714 erroneous result when trying to check for a repeat of the original
2715 character.
2717 13. Some tidying changes to the testing mechanism:
2719 (a) The RunTest script now detects the internal link size and whether there
2720 is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2721 values substituted by "configure". (The RunGrepTest script already did
2722 this for UTF-8.) The configure.ac script no longer substitutes the
2723 relevant variables.
2725 (b) The debugging options /B and /D in pcretest show the compiled bytecode
2726 with length and offset values. This means that the output is different
2727 for different internal link sizes. Test 2 is skipped for link sizes
2728 other than 2 because of this, bypassing the problem. Unfortunately,
2729 there was also a test in test 3 (the locale tests) that used /B and
2730 failed for link sizes other than 2. Rather than cut the whole test out,
2731 I have added a new /Z option to pcretest that replaces the length and
2732 offset values with spaces. This is now used to make test 3 independent
2733 of link size. (Test 2 will be tidied up later.)
2735 14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2736 segmentation fault instead of returning the appropriate error message.
2738 15. In multiline mode when the newline sequence was set to "any", the pattern
2739 ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2740 This doesn't seem right; it now treats the CRLF combination as the line
2741 ending, and so does not match in that case. It's only a pattern such as ^$
2742 that would hit this one: something like ^ABC$ would have failed after \r
2743 and then tried again after \r\n.
2745 16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2746 in an attempt to make files that differ only in their line terminators
2747 compare equal. This works on Linux.
2749 17. Under certain error circumstances pcregrep might try to free random memory
2750 as it exited. This is now fixed, thanks to valgrind.
2752 19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2753 "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2754 was because its rules for how to advance for /g after matching an empty
2755 string at the end of a line did not allow for this case. They now check for
2756 it specially.
2758 20. pcretest is supposed to handle patterns and data of any length, by
2759 extending its buffers when necessary. It was getting this wrong when the
2760 buffer for a data line had to be extended.
2762 21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2763 CRLF as a newline sequence.
2765 22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2766 out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2767 I have nevertheless tidied it up.
2769 23. Added some casts to kill warnings from HP-UX ia64 compiler.
2771 24. Added a man page for pcre-config.
2774 Version 7.0 19-Dec-06
2775 ---------------------
2777 1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2778 moving to gcc 4.1.1.
2780 2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2781 sys/time.h, which is documented as needed for this function. It doesn't
2782 seem to matter on Linux, but it showed up on some releases of OS X.
2784 3. It seems that there are systems where bytes whose values are greater than
2785 127 match isprint() in the "C" locale. The "C" locale should be the
2786 default when a C program starts up. In most systems, only ASCII printing
2787 characters match isprint(). This difference caused the output from pcretest
2788 to vary, making some of the tests fail. I have changed pcretest so that:
2790 (a) When it is outputting text in the compiled version of a pattern, bytes
2791 other than 32-126 are always shown as hex escapes.
2793 (b) When it is outputting text that is a matched part of a subject string,
2794 it does the same, unless a different locale has been set for the match
2795 (using the /L modifier). In this case, it uses isprint() to decide.
2797 4. Fixed a major bug that caused incorrect computation of the amount of memory
2798 required for a compiled pattern when options that changed within the
2799 pattern affected the logic of the preliminary scan that determines the
2800 length. The relevant options are -x, and -i in UTF-8 mode. The result was
2801 that the computed length was too small. The symptoms of this bug were
2802 either the PCRE error "internal error: code overflow" from pcre_compile(),
2803 or a glibc crash with a message such as "pcretest: free(): invalid next
2804 size (fast)". Examples of patterns that provoked this bug (shown in
2805 pcretest format) are:
2807 /(?-x: )/x
2808 /(?x)(?-x: \s*#\s*)/
2809 /((?i)[\x{c0}])/8
2810 /(?i:[\x{c0}])/8
2812 HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2813 is now done differently.
2815 5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2816 wrapper classes; (b) implement a new function in the C++ scanner that is
2817 more efficient than the old way of doing things because it avoids levels of
2818 recursion in the regex matching; (c) add a paragraph to the documentation
2819 for the FullMatch() function.
2821 6. The escape sequence \n was being treated as whatever was defined as
2822 "newline". Not only was this contrary to the documentation, which states
2823 that \n is character 10 (hex 0A), but it also went horribly wrong when
2824 "newline" was defined as CRLF. This has been fixed.
2826 7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2827 was being set to -1 for the "end of line" case (supposedly a value that no
2828 character can have). Though this value is never used (the check for end of
2829 line is "zero bytes in current character"), it caused compiler complaints.
2830 I've changed it to 0xffffffff.
2832 8. In pcre_version.c, the version string was being built by a sequence of
2833 C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2834 string (as it is for production releases) called a macro with an empty
2835 argument. The C standard says the result of this is undefined. The gcc
2836 compiler treats it as an empty string (which was what was wanted) but it is
2837 reported that Visual C gives an error. The source has been hacked around to
2838 avoid this problem.
2840 9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2841 builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2842 instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2843 of them did).
2845 10. Originally, pcretest opened its input and output without "b"; then I was
2846 told that "b" was needed in some environments, so it was added for release
2847 5.0 to both the input and output. (It makes no difference on Unix-like
2848 systems.) Later I was told that it is wrong for the input on Windows. I've
2849 now abstracted the modes into two macros, to make it easier to fiddle with
2850 them, and removed "b" from the input mode under Windows.
2852 11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2854 12. Added -help and --help to pcretest as an official way of being reminded
2855 of the options.
2857 13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2858 and pcrecpp.cc because they annoy compilers at high warning levels.
2860 14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2862 15. Fixed an occurrence of == in configure.ac that should have been = (shell
2863 scripts are not C programs :-) and which was not noticed because it works
2864 on Linux.
2866 16. pcretest is supposed to handle any length of pattern and data line (as one
2867 line or as a continued sequence of lines) by extending its input buffer if
2868 necessary. This feature was broken for very long pattern lines, leading to
2869 a string of junk being passed to pcre_compile() if the pattern was longer
2870 than about 50K.
2872 17. I have done a major re-factoring of the way pcre_compile() computes the
2873 amount of memory needed for a compiled pattern. Previously, there was code
2874 that made a preliminary scan of the pattern in order to do this. That was
2875 OK when PCRE was new, but as the facilities have expanded, it has become
2876 harder and harder to keep it in step with the real compile phase, and there
2877 have been a number of bugs (see for example, 4 above). I have now found a
2878 cunning way of running the real compile function in a "fake" mode that
2879 enables it to compute how much memory it would need, while actually only
2880 ever using a few hundred bytes of working memory and without too many
2881 tests of the mode. This should make future maintenance and development
2882 easier. A side effect of this work is that the limit of 200 on the nesting
2883 depth of parentheses has been removed (though this was never a serious
2884 limitation, I suspect). However, there is a downside: pcre_compile() now
2885 runs more slowly than before (30% or more, depending on the pattern). I
2886 hope this isn't a big issue. There is no effect on runtime performance.
2888 18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2889 newline (only possible for the last line of a file) and it was a
2890 pattern that set a locale (followed by /Lsomething), pcretest crashed.
2892 19. Added additional timing features to pcretest. (1) The -tm option now times
2893 matching only, not compiling. (2) Both -t and -tm can be followed, as a
2894 separate command line item, by a number that specifies the number of
2895 repeats to use when timing. The default is 50000; this gives better
2896 precision, but takes uncomfortably long for very large patterns.
2898 20. Extended pcre_study() to be more clever in cases where a branch of a
2899 subpattern has no definite first character. For example, (a*|b*)[cd] would
2900 previously give no result from pcre_study(). Now it recognizes that the
2901 first character must be a, b, c, or d.
2903 21. There was an incorrect error "recursive call could loop indefinitely" if
2904 a subpattern (or the entire pattern) that was being tested for matching an
2905 empty string contained only one non-empty item after a nested subpattern.
2906 For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2907 incorrectly, because the \d was being skipped in the check.
2909 22. The pcretest program now has a new pattern option /B and a command line
2910 option -b, which is equivalent to adding /B to every pattern. This causes
2911 it to show the compiled bytecode, without the additional information that
2912 -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2913 is the same as /B/I).
2915 23. A new optimization is now able automatically to treat some sequences such
2916 as a*b as a*+b. More specifically, if something simple (such as a character
2917 or a simple class like \d) has an unlimited quantifier, and is followed by
2918 something that cannot possibly match the quantified thing, the quantifier
2919 is automatically "possessified".
2921 24. A recursive reference to a subpattern whose number was greater than 39
2922 went wrong under certain circumstances in UTF-8 mode. This bug could also
2923 have affected the operation of pcre_study().
2925 25. Realized that a little bit of performance could be had by replacing
2926 (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
2928 26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
2930 27. Possessive quantifiers such as a++ were previously implemented by turning
2931 them into atomic groups such as ($>a+). Now they have their own opcodes,
2932 which improves performance. This includes the automatically created ones
2933 from 23 above.
2935 28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
2936 lookahead was broken if it was not anchored. PCRE was mistakenly expecting
2937 the first matched character to be a colon. This applied both to named and
2938 numbered groups.
2940 29. The ucpinternal.h header file was missing its idempotency #ifdef.
2942 30. I was sent a "project" file called libpcre.a.dev which I understand makes
2943 building PCRE on Windows easier, so I have included it in the distribution.
2945 31. There is now a check in pcretest against a ridiculously large number being
2946 returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
2947 loop, the loop is abandoned.
2949 32. Forward references to subpatterns in conditions such as (?(2)...) where
2950 subpattern 2 is defined later cause pcre_compile() to search forwards in
2951 the pattern for the relevant set of parentheses. This search went wrong
2952 when there were unescaped parentheses in a character class, parentheses
2953 escaped with \Q...\E, or parentheses in a #-comment in /x mode.
2955 33. "Subroutine" calls and backreferences were previously restricted to
2956 referencing subpatterns earlier in the regex. This restriction has now
2957 been removed.
2959 34. Added a number of extra features that are going to be in Perl 5.10. On the
2960 whole, these are just syntactic alternatives for features that PCRE had
2961 previously implemented using the Python syntax or my own invention. The
2962 other formats are all retained for compatibility.
2964 (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
2965 as (?P<name>...). The new forms, as well as being in Perl 5.10, are
2966 also .NET compatible.
2968 (b) A recursion or subroutine call to a named group can now be defined as
2969 (?&name) as well as (?P>name).
2971 (c) A backreference to a named group can now be defined as \k<name> or
2972 \k'name' as well as (?P=name). The new forms, as well as being in Perl
2973 5.10, are also .NET compatible.
2975 (d) A conditional reference to a named group can now use the syntax
2976 (?(<name>) or (?('name') as well as (?(name).
2978 (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
2979 groups (named and numbered) that are never evaluated inline, but can be
2980 called as "subroutines" from elsewhere. In effect, the DEFINE condition
2981 is always false. There may be only one alternative in such a group.
2983 (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
2984 as the simple (?(R). The condition is true only if the most recent
2985 recursion is that of the given number or name. It does not search out
2986 through the entire recursion stack.
2988 (g) The escape \gN or \g{N} has been added, where N is a positive or
2989 negative number, specifying an absolute or relative reference.
2991 35. Tidied to get rid of some further signed/unsigned compiler warnings and
2992 some "unreachable code" warnings.
2994 36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
2995 things, this adds five new scripts.
2997 37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
2998 There were also incompatibilities regarding the handling of \Q..\E inside
2999 character classes, for example with patterns like [\Qa\E-\Qz\E] where the
3000 hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
3002 38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
3003 matches an empty string, and forcibly breaks the loop. There were bugs in
3004 this code in non-simple cases. For a pattern such as ^(a()*)* matched
3005 against aaaa the result was just "a" rather than "aaaa", for example. Two
3006 separate and independent bugs (that affected different cases) have been
3007 fixed.
3009 39. Refactored the code to abolish the use of different opcodes for small
3010 capturing bracket numbers. This is a tidy that I avoided doing when I
3011 removed the limit on the number of capturing brackets for 3.5 back in 2001.
3012 The new approach is not only tidier, it makes it possible to reduce the
3013 memory needed to fix the previous bug (38).
3015 40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3016 sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3017 processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3018 mode.
3020 41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3021 report.
3023 42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3024 copy construction and assignment in the C++ wrapper.
3026 43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3027 couple of bugs that could have given wrong results in the "--newline=crlf"
3028 case.
3030 44. Added a number of casts and did some reorganization of signed/unsigned int
3031 variables following suggestions from Dair Grant. Also renamed the variable
3032 "this" as "item" because it is a C++ keyword.
3034 45. Arranged for dftables to add
3036 #include "pcre_internal.h"
3038 to pcre_chartables.c because without it, gcc 4.x may remove the array
3039 definition from the final binary if PCRE is built into a static library and
3040 dead code stripping is activated.
3042 46. For an unanchored pattern, if a match attempt fails at the start of a
3043 newline sequence, and the newline setting is CRLF or ANY, and the next two
3044 characters are CRLF, advance by two characters instead of one.
3047 Version 6.7 04-Jul-06
3048 ---------------------
3050 1. In order to handle tests when input lines are enormously long, pcretest has
3051 been re-factored so that it automatically extends its buffers when
3052 necessary. The code is crude, but this _is_ just a test program. The
3053 default size has been increased from 32K to 50K.
3055 2. The code in pcre_study() was using the value of the re argument before
3056 testing it for NULL. (Of course, in any sensible call of the function, it
3057 won't be NULL.)
3059 3. The memmove() emulation function in pcre_internal.h, which is used on
3060 systems that lack both memmove() and bcopy() - that is, hardly ever -
3061 was missing a "static" storage class specifier.
3063 4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3064 containing an extended class (one that cannot be represented by a bitmap
3065 because it contains high-valued characters or Unicode property items, e.g.
3066 [\pZ]). Almost always one would set UTF-8 mode when processing such a
3067 pattern, but PCRE should not loop if you do not (it no longer does).
3068 [Detail: two cases were found: (a) a repeated subpattern containing an
3069 extended class; (b) a recursive reference to a subpattern that followed a
3070 previous extended class. It wasn't skipping over the extended class
3071 correctly when UTF-8 mode was not set.]
3073 5. A negated single-character class was not being recognized as fixed-length
3074 in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3075 compile error "lookbehind assertion is not fixed length".
3077 6. The RunPerlTest auxiliary script was showing an unexpected difference
3078 between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3079 write a Perl script that can interpret lines of an input file either as
3080 byte characters or as UTF-8, which is what "perltest" was being required to
3081 do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3082 can't do is switch easily at run time between having the "use utf8;" pragma
3083 or not. In the end, I fudged it by using the RunPerlTest script to insert
3084 "use utf8;" explicitly for the UTF-8 tests.
3086 7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3087 the end of the subject string, contrary to the documentation and to what
3088 Perl does. This was true of both matching functions. Now it matches only at
3089 the start of the subject and immediately after *internal* newlines.
3091 8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3092 a pointer to an int instead of a pointer to an unsigned long int. This
3093 caused problems on 64-bit systems.
3095 9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3096 instance of the 'standard' template library not being so standard".
3098 10. There was no check on the number of named subpatterns nor the maximum
3099 length of a subpattern name. The product of these values is used to compute
3100 the size of the memory block for a compiled pattern. By supplying a very
3101 long subpattern name and a large number of named subpatterns, the size
3102 computation could be caused to overflow. This is now prevented by limiting
3103 the length of names to 32 characters, and the number of named subpatterns
3104 to 10,000.
3106 11. Subpatterns that are repeated with specific counts have to be replicated in
3107 the compiled pattern. The size of memory for this was computed from the
3108 length of the subpattern and the repeat count. The latter is limited to
3109 65535, but there was no limit on the former, meaning that integer overflow
3110 could in principle occur. The compiled length of a repeated subpattern is
3111 now limited to 30,000 bytes in order to prevent this.
3113 12. Added the optional facility to have named substrings with the same name.
3115 13. Added the ability to use a named substring as a condition, using the
3116 Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3117 are numbers (not recommended). Forward references are permitted.
3119 14. Added forward references in named backreferences (if you see what I mean).
3121 15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3122 pattern could run off the end of the subject. For example, the pattern
3123 "(?s)(.{1,5})"8 did this with the subject "ab".
3125 16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3126 PCRE_CASELESS was set when matching characters that were quantified with ?
3127 or *.
3129 17. A character class other than a single negated character that had a minimum
3130 but no maximum quantifier - for example [ab]{6,} - was not handled
3131 correctly by pce_dfa_exec(). It would match only one character.
3133 18. A valid (though odd) pattern that looked like a POSIX character
3134 class but used an invalid character after [ (for example [[,abc,]]) caused
3135 pcre_compile() to give the error "Failed: internal error: code overflow" or
3136 in some cases to crash with a glibc free() error. This could even happen if
3137 the pattern terminated after [[ but there just happened to be a sequence of
3138 letters, a binary zero, and a closing ] in the memory that followed.
3140 19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3141 over the years. Originally (before any Unicode support), just the bottom 8
3142 bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3143 output from "man perlunicode" includes this:
3145 The regular expression compiler produces polymorphic opcodes. That
3146 is, the pattern adapts to the data and automatically switches to
3147 the Unicode character scheme when presented with Unicode data--or
3148 instead uses a traditional byte scheme when presented with byte
3149 data.
3151 Sadly, a wide octal escape does not cause a switch, and in a string with
3152 no other multibyte characters, these octal escapes are treated as before.
3153 Thus, in Perl, the pattern /\500/ actually matches \100 but the pattern
3154 /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3155 Unicode string.
3157 I have not perpetrated such confusion in PCRE. Up till now, it took just
3158 the bottom 8 bits, as in old Perl. I have now made octal escapes with
3159 values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3160 translate to the appropriate multibyte character.
3162 29. Applied some refactoring to reduce the number of warnings from Microsoft
3163 and Borland compilers. This has included removing the fudge introduced
3164 seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3165 a warning about an unused variable.
3167 21. PCRE has not included VT (character 0x0b) in the set of whitespace
3168 characters since release 4.0, because Perl (from release 5.004) does not.
3169 [Or at least, is documented not to: some releases seem to be in conflict
3170 with the documentation.] However, when a pattern was studied with
3171 pcre_study() and all its branches started with \s, PCRE still included VT
3172 as a possible starting character. Of course, this did no harm; it just
3173 caused an unnecessary match attempt.
3175 22. Removed a now-redundant internal flag bit that recorded the fact that case
3176 dependency changed within the pattern. This was once needed for "required
3177 byte" processing, but is no longer used. This recovers a now-scarce options
3178 bit. Also moved the least significant internal flag bit to the most-
3179 significant bit of the word, which was not previously used (hangover from
3180 the days when it was an int rather than a uint) to free up another bit for
3181 the future.
3183 23. Added support for CRLF line endings as well as CR and LF. As well as the
3184 default being selectable at build time, it can now be changed at runtime
3185 via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3186 specify that it is scanning data with non-default line endings.
3188 24. Changed the definition of CXXLINK to make it agree with the definition of
3189 LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3191 25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3192 recursions. This makes a big different to stack usage for some patterns.
3194 26. If a subpattern containing a named recursion or subroutine reference such
3195 as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3196 the space required for the compiled pattern went wrong and gave too small a
3197 value. Depending on the environment, this could lead to "Failed: internal
3198 error: code overflow at offset 49" or "glibc detected double free or
3199 corruption" errors.
3201 27. Applied patches from Google (a) to support the new newline modes and (b) to
3202 advance over multibyte UTF-8 characters in GlobalReplace.
3204 28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3205 difference for some implementation of PCRE in some Windows version.
3207 29. Added some extra testing facilities to pcretest:
3209 \q<number> in a data line sets the "match limit" value
3210 \Q<number> in a data line sets the "match recursion limt" value
3211 -S <number> sets the stack size, where <number> is in megabytes
3213 The -S option isn't available for Windows.
3216 Version 6.6 06-Feb-06
3217 ---------------------
3219 1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3220 in pcreposix.h. I have copied the definition from pcre.h.
3222 2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3223 because pcre.h is no longer a built file.
3225 3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3226 not normally included in the compiled code.
3229 Version 6.5 01-Feb-06
3230 ---------------------
3232 1. When using the partial match feature with pcre_dfa_exec(), it was not
3233 anchoring the second and subsequent partial matches at the new starting
3234 point. This could lead to incorrect results. For example, with the pattern
3235 /1234/, partially matching against "123" and then "a4" gave a match.
3237 2. Changes to pcregrep:
3239 (a) All non-match returns from pcre_exec() were being treated as failures
3240 to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3241 error message is output. Some extra information is given for the
3243 probably the only errors that are likely to be caused by users (by
3244 specifying a regex that has nested indefinite repeats, for instance).
3245 If there are more than 20 of these errors, pcregrep is abandoned.
3247 (b) A binary zero was treated as data while matching, but terminated the
3248 output line if it was written out. This has been fixed: binary zeroes
3249 are now no different to any other data bytes.
3251 (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3252 used to set a locale for matching. The --locale=xxxx long option has
3253 been added (no short equivalent) to specify a locale explicitly on the
3254 pcregrep command, overriding the environment variables.
3256 (d) When -B was used with -n, some line numbers in the output were one less
3257 than they should have been.
3259 (e) Added the -o (--only-matching) option.
3261 (f) If -A or -C was used with -c (count only), some lines of context were
3262 accidentally printed for the final match.
3264 (g) Added the -H (--with-filename) option.
3266 (h) The combination of options -rh failed to suppress file names for files
3267 that were found from directory arguments.
3269 (i) Added the -D (--devices) and -d (--directories) options.
3271 (j) Added the -F (--fixed-strings) option.
3273 (k) Allow "-" to be used as a file name for -f as well as for a data file.
3275 (l) Added the --colo(u)r option.
3277 (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3278 is not present by default.
3280 3. A nasty bug was discovered in the handling of recursive patterns, that is,
3281 items such as (?R) or (?1), when the recursion could match a number of
3282 alternatives. If it matched one of the alternatives, but subsequently,
3283 outside the recursion, there was a failure, the code tried to back up into
3284 the recursion. However, because of the way PCRE is implemented, this is not
3285 possible, and the result was an incorrect result from the match.
3287 In order to prevent this happening, the specification of recursion has
3288 been changed so that all such subpatterns are automatically treated as
3289 atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3291 4. I had overlooked the fact that, in some locales, there are characters for
3292 which isalpha() is true but neither isupper() nor islower() are true. In
3293 the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3294 and ordfeminine) are like this. This affected the treatment of \w and \W
3295 when they appeared in character classes, but not when they appeared outside
3296 a character class. The bit map for "word" characters is now created
3297 separately from the results of isalnum() instead of just taking it from the
3298 upper, lower, and digit maps. (Plus the underscore character, of course.)
3300 5. The above bug also affected the handling of POSIX character classes such as
3301 [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3302 permanent tables. Instead, the bit maps for such a class were previously
3303 created as the appropriate unions of the upper, lower, and digit bitmaps.
3304 Now they are created by subtraction from the [[:word:]] class, which has
3305 its own bitmap.
3307 6. The [[:blank:]] character class matches horizontal, but not vertical space.
3308 It is created by subtracting the vertical space characters (\x09, \x0a,
3309 \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3310 subtraction was done in the overall bitmap for a character class, meaning
3311 that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3312 be recognized. This bug has been fixed.
3314 7. Patches from the folks at Google:
3316 (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3317 real life, but is still worth protecting against".
3319 (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3320 regular expressions".
3322 (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3323 have it.
3325 (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3326 "configure" and the latter not, in order to fix a problem somebody had
3327 with compiling the Arg class on HP-UX.
3329 (e) Improve the error-handling of the C++ wrapper a little bit.
3331 (f) New tests for checking recursion limiting.
3333 8. The pcre_memmove() function, which is used only if the environment does not
3334 have a standard memmove() function (and is therefore rarely compiled),
3335 contained two bugs: (a) use of int instead of size_t, and (b) it was not
3336 returning a result (though PCRE never actually uses the result).
3338 9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3339 large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3340 returned instead of calling malloc() with an overflowing number that would
3341 most likely cause subsequent chaos.
3343 10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3345 11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3346 with this option is matched, the nmatch and pmatch options of regexec() are
3347 ignored.
3349 12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3350 provided in case anyone wants to the the POSIX interface with UTF-8
3351 strings.
3353 13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3354 C++ linking (needed for some HP-UX environments).
3356 14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3357 (unused parameter) and in the pcre_printint() function (omitted "default"
3358 switch label when the default is to do nothing).
3360 15. Added some code to make it possible, when PCRE is compiled as a C++
3361 library, to replace subject pointers for pcre_exec() with a smart pointer
3362 class, thus making it possible to process discontinuous strings.
3364 16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3365 much the same function. They were added by different people who were trying
3366 to make PCRE easy to compile on non-Unix systems. It has been suggested
3367 that PCRE_EXPORT be abolished now that there is more automatic apparatus
3368 for compiling on Windows systems. I have therefore replaced it with
3369 PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3370 defaults to "extern" for C or "extern C" for C++, which works fine on
3371 Unix-like systems. It is now possible to override the value of PCRE_DATA_
3372 SCOPE with something explicit in config.h. In addition:
3374 (a) pcreposix.h still had just "extern" instead of either of these macros;
3375 I have replaced it with PCRE_DATA_SCOPE.
3377 (b) Functions such as _pcre_xclass(), which are internal to the library,
3378 but external in the C sense, all had PCRE_EXPORT in their definitions.
3379 This is apparently wrong for the Windows case, so I have removed it.
3380 (It makes no difference on Unix-like systems.)
3382 17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3383 of recursive calls to match(). This is different to MATCH_LIMIT because
3384 that limits the total number of calls to match(), not all of which increase
3385 the depth of recursion. Limiting the recursion depth limits the amount of
3386 stack (or heap if NO_RECURSE is set) that is used. The default can be set
3387 when PCRE is compiled, and changed at run time. A patch from Google adds
3388 this functionality to the C++ interface.
3390 18. Changes to the handling of Unicode character properties:
3392 (a) Updated the table to Unicode 4.1.0.
3394 (b) Recognize characters that are not in the table as "Cn" (undefined).
3396 (c) I revised the way the table is implemented to a much improved format
3397 which includes recognition of ranges. It now supports the ranges that
3398 are defined in UnicodeData.txt, and it also amalgamates other
3399 characters into ranges. This has reduced the number of entries in the
3400 table from around 16,000 to around 3,000, thus reducing its size
3401 considerably. I realized I did not need to use a tree structure after
3402 all - a binary chop search is just as efficient. Having reduced the
3403 number of entries, I extended their size from 6 bytes to 8 bytes to
3404 allow for more data.
3406 (d) Added support for Unicode script names via properties such as \p{Han}.
3408 19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3409 matching that character.
3411 20. When matching a repeated Unicode property with a minimum greater than zero,
3412 (for example \pL{2,}), PCRE could look past the end of the subject if it
3413 reached it while seeking the minimum number of characters. This could
3414 happen only if some of the characters were more than one byte long, because
3415 there is a check for at least the minimum number of bytes.
3417 21. Refactored the implementation of \p and \P so as to be more general, to
3418 allow for more different types of property in future. This has changed the
3419 compiled form incompatibly. Anybody with saved compiled patterns that use
3420 \p or \P will have to recompile them.
3422 22. Added "Any" and "L&" to the supported property types.
3424 23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3425 but give a compile time error if the value is greater than 0xff.
3427 24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3428 accidentally not being installed or uninstalled.
3430 25. The pcre.h file was built from pcre.h.in, but the only changes that were
3431 made were to insert the current release number. This seemed silly, because
3432 it made things harder for people building PCRE on systems that don't run
3433 "configure". I have turned pcre.h into a distributed file, no longer built
3434 by "configure", with the version identification directly included. There is
3435 no longer a pcre.h.in file.
3437 However, this change necessitated a change to the pcre-config script as
3438 well. It is built from pcre-config.in, and one of the substitutions was the
3439 release number. I have updated configure.ac so that ./configure now finds
3440 the release number by grepping pcre.h.
3442 26. Added the ability to run the tests under valgrind.
3445 Version 6.4 05-Sep-05
3446 ---------------------
3448 1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3449 "--" to be printed when multiple files were scanned, even when none of the
3450 -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3451 consider it to be a bug, and have restored the previous behaviour.
3453 2. A couple of code tidies to get rid of compiler warnings.
3455 3. The pcretest program used to cheat by referring to symbols in the library
3456 whose names begin with _pcre_. These are internal symbols that are not
3457 really supposed to be visible externally, and in some environments it is
3458 possible to suppress them. The cheating is now confined to including
3459 certain files from the library's source, which is a bit cleaner.
3461 4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3462 file's purpose clearer.
3464 5. Reorganized pcre_ucp_findchar().
3467 Version 6.3 15-Aug-05
3468 ---------------------
3470 1. The file libpcre.pc.in did not have general read permission in the tarball.
3472 2. There were some problems when building without C++ support:
3474 (a) If C++ support was not built, "make install" and "make test" still
3475 tried to test it.
3477 (b) There were problems when the value of CXX was explicitly set. Some
3478 changes have been made to try to fix these, and ...
3480 (c) --disable-cpp can now be used to explicitly disable C++ support.
3482 (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3483 backslash in a target when C++ was disabled. This confuses some
3484 versions of "make", apparently. Using an intermediate variable solves
3485 this. (Same for CPP_LOBJ.)
3487 3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3488 (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3489 necessary on certain architectures.
3491 4. Added a setting of -export-symbols-regex to the link command to remove
3492 those symbols that are exported in the C sense, but actually are local
3493 within the library, and not documented. Their names all begin with
3494 "_pcre_". This is not a perfect job, because (a) we have to except some
3495 symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3496 available (and never for static libraries). I have made a note to try to
3497 find a way round (a) in the future.
3500 Version 6.2 01-Aug-05
3501 ---------------------
3503 1. There was no test for integer overflow of quantifier values. A construction
3504 such as {1111111111111111} would give undefined results. What is worse, if
3505 a minimum quantifier for a parenthesized subpattern overflowed and became
3506 negative, the calculation of the memory size went wrong. This could have
3507 led to memory overwriting.
3509 2. Building PCRE using VPATH was broken. Hopefully it is now fixed.
3511 3. Added "b" to the 2nd argument of fopen() in dftables.c, for non-Unix-like
3512 operating environments where this matters.
3514 4. Applied Giuseppe Maxia's patch to add additional features for controlling
3515 PCRE options from within the C++ wrapper.
3517 5. Named capturing subpatterns were not being correctly counted when a pattern
3518 was compiled. This caused two problems: (a) If there were more than 100
3519 such subpatterns, the calculation of the memory needed for the whole
3520 compiled pattern went wrong, leading to an overflow error. (b) Numerical
3521 back references of the form \12, where the number was greater than 9, were
3522 not recognized as back references, even though there were sufficient
3523 previous subpatterns.
3525 6. Two minor patches to pcrecpp.cc in order to allow it to compile on older
3526 versions of gcc, e.g. 2.95.4.
3529 Version 6.1 21-Jun-05
3530 ---------------------
3532 1. There was one reference to the variable "posix" in pcretest.c that was not
3533 surrounded by "#if !defined NOPOSIX".
3535 2. Make it possible to compile pcretest without DFA support, UTF8 support, or
3536 the cross-check on the old pcre_info() function, for the benefit of the
3537 cut-down version of PCRE that is currently imported into Exim.
3539 3. A (silly) pattern starting with (?i)(?-i) caused an internal space
3540 allocation error. I've done the easy fix, which wastes 2 bytes for sensible
3541 patterns that start (?i) but I don't think that matters. The use of (?i) is
3542 just an example; this all applies to the other options as well.
3544 4. Since libtool seems to echo the compile commands it is issuing, the output
3545 from "make" can be reduced a bit by putting "@" in front of each libtool
3546 compile command.
3548 5. Patch from the folks at Google for configure.in to be a bit more thorough
3549 in checking for a suitable C++ installation before trying to compile the
3550 C++ stuff. This should fix a reported problem when a compiler was present,
3551 but no suitable headers.
3553 6. The man pages all had just "PCRE" as their title. I have changed them to
3554 be the relevant file name. I have also arranged that these names are
3555 retained in the file doc/pcre.txt, which is a concatenation in text format
3556 of all the man pages except the little individual ones for each function.
3558 7. The NON-UNIX-USE file had not been updated for the different set of source
3559 files that come with release 6. I also added a few comments about the C++
3560 wrapper.
3563 Version 6.0 07-Jun-05
3564 ---------------------
3566 1. Some minor internal re-organization to help with my DFA experiments.
3568 2. Some missing #ifdef SUPPORT_UCP conditionals in pcretest and printint that
3569 didn't matter for the library itself when fully configured, but did matter
3570 when compiling without UCP support, or within Exim, where the ucp files are
3571 not imported.
3573 3. Refactoring of the library code to split up the various functions into
3574 different source modules. The addition of the new DFA matching code (see
3575 below) to a single monolithic source would have made it really too
3576 unwieldy, quite apart from causing all the code to be include in a
3577 statically linked application, when only some functions are used. This is
3578 relevant even without the DFA addition now that patterns can be compiled in
3579 one application and matched in another.
3581 The downside of splitting up is that there have to be some external
3582 functions and data tables that are used internally in different modules of
3583 the library but which are not part of the API. These have all had their
3584 names changed to start with "_pcre_" so that they are unlikely to clash
3585 with other external names.
3587 4. Added an alternate matching function, pcre_dfa_exec(), which matches using
3588 a different (DFA) algorithm. Although it is slower than the original
3589 function, it does have some advantages for certain types of matching
3590 problem.
3592 5. Upgrades to pcretest in order to test the features of pcre_dfa_exec(),
3593 including restarting after a partial match.
3595 6. A patch for pcregrep that defines INVALID_FILE_ATTRIBUTES if it is not
3596 defined when compiling for Windows was sent to me. I have put it into the
3597 code, though I have no means of testing or verifying it.
3599 7. Added the pcre_refcount() auxiliary function.
3601 8. Added the PCRE_FIRSTLINE option. This constrains an unanchored pattern to
3602 match before or at the first newline in the subject string. In pcretest,
3603 the /f option on a pattern can be used to set this.
3605 9. A repeated \w when used in UTF-8 mode with characters greater than 256
3606 would behave wrongly. This has been present in PCRE since release 4.0.
3608 10. A number of changes to the pcregrep command:
3610 (a) Refactored how -x works; insert ^(...)$ instead of setting
3611 PCRE_ANCHORED and checking the length, in preparation for adding
3612 something similar for -w.
3614 (b) Added the -w (match as a word) option.
3616 (c) Refactored the way lines are read and buffered so as to have more
3617 than one at a time available.
3619 (d) Implemented a pcregrep test script.
3621 (e) Added the -M (multiline match) option. This allows patterns to match
3622 over several lines of the subject. The buffering ensures that at least
3623 8K, or the rest of the document (whichever is the shorter) is available
3624 for matching (and similarly the previous 8K for lookbehind assertions).
3626 (f) Changed the --help output so that it now says
3628 -w, --word-regex(p)
3630 instead of two lines, one with "regex" and the other with "regexp"
3631 because that confused at least one person since the short forms are the
3632 same. (This required a bit of code, as the output is generated
3633 automatically from a table. It wasn't just a text change.)
3635 (g) -- can be used to terminate pcregrep options if the next thing isn't an
3636 option but starts with a hyphen. Could be a pattern or a path name
3637 starting with a hyphen, for instance.
3639 (h) "-" can be given as a file name to represent stdin.
3641 (i) When file names are being printed, "(standard input)" is used for
3642 the standard input, for compatibility with GNU grep. Previously
3643 "<stdin>" was used.
3645 (j) The option --label=xxx can be used to supply a name to be used for
3646 stdin when file names are being printed. There is no short form.
3648 (k) Re-factored the options decoding logic because we are going to add
3649 two more options that take data. Such options can now be given in four
3650 different ways, e.g. "-fname", "-f name", "--file=name", "--file name".
3652 (l) Added the -A, -B, and -C options for requesting that lines of context
3653 around matches be printed.
3655 (m) Added the -L option to print the names of files that do not contain
3656 any matching lines, that is, the complement of -l.
3658 (n) The return code is 2 if any file cannot be opened, but pcregrep does
3659 continue to scan other files.
3661 (o) The -s option was incorrectly implemented. For compatibility with other
3662 greps, it now suppresses the error message for a non-existent or non-
3663 accessible file (but not the return code). There is a new option called
3664 -q that suppresses the output of matching lines, which was what -s was
3665 previously doing.
3667 (p) Added --include and --exclude options to specify files for inclusion
3668 and exclusion when recursing.
3670 11. The Makefile was not using the Autoconf-supported LDFLAGS macro properly.
3671 Hopefully, it now does.
3673 12. Missing cast in pcre_study().
3675 13. Added an "uninstall" target to the makefile.
3677 14. Replaced "extern" in the function prototypes in Makefile.in with
3678 "PCRE_DATA_SCOPE", which defaults to 'extern' or 'extern "C"' in the Unix
3679 world, but is set differently for Windows.
3681 15. Added a second compiling function called pcre_compile2(). The only
3682 difference is that it has an extra argument, which is a pointer to an
3683 integer error code. When there is a compile-time failure, this is set
3684 non-zero, in addition to the error test pointer being set to point to an
3685 error message. The new argument may be NULL if no error number is required
3686 (but then you may as well call pcre_compile(), which is now just a
3687 wrapper). This facility is provided because some applications need a
3688 numeric error indication, but it has also enabled me to tidy up the way
3689 compile-time errors are handled in the POSIX wrapper.
3691 16. Added VPATH=.libs to the makefile; this should help when building with one
3692 prefix path and installing with another. (Or so I'm told by someone who
3693 knows more about this stuff than I do.)
3695 17. Added a new option, REG_DOTALL, to the POSIX function regcomp(). This
3696 passes PCRE_DOTALL to the pcre_compile() function, making the "." character
3697 match everything, including newlines. This is not POSIX-compatible, but
3698 somebody wanted the feature. From pcretest it can be activated by using
3699 both the P and the s flags.
3701 18. AC_PROG_LIBTOOL appeared twice in Makefile.in. Removed one.
3703 19. libpcre.pc was being incorrectly installed as executable.
3705 20. A couple of places in pcretest check for end-of-line by looking for '\n';
3706 it now also looks for '\r' so that it will work unmodified on Windows.
3708 21. Added Google's contributed C++ wrapper to the distribution.
3710 22. Added some untidy missing memory free() calls in pcretest, to keep
3711 Electric Fence happy when testing.
3715 Version 5.0 13-Sep-04
3716 ---------------------
3718 1. Internal change: literal characters are no longer packed up into items
3719 containing multiple characters in a single byte-string. Each character
3720 is now matched using a separate opcode. However, there may be more than one
3721 byte in the character in UTF-8 mode.
3723 2. The pcre_callout_block structure has two new fields: pattern_position and
3724 next_item_length. These contain the offset in the pattern to the next match
3725 item, and its length, respectively.
3727 3. The PCRE_AUTO_CALLOUT option for pcre_compile() requests the automatic
3728 insertion of callouts before each pattern item. Added the /C option to
3729 pcretest to make use of this.
3731 4. On the advice of a Windows user, the lines
3733 #if defined(_WIN32) || defined(WIN32)
3734 _setmode( _fileno( stdout ), 0x8000 );
3735 #endif /* defined(_WIN32) || defined(WIN32) */
3737 have been added to the source of pcretest. This apparently does useful
3738 magic in relation to line terminators.
3740 5. Changed "r" and "w" in the calls to fopen() in pcretest to "rb" and "wb"
3741 for the benefit of those environments where the "b" makes a difference.
3743 6. The icc compiler has the same options as gcc, but "configure" doesn't seem
3744 to know about it. I have put a hack into configure.in that adds in code
3745 to set GCC=yes if CC=icc. This seems to end up at a point in the
3746 generated configure script that is early enough to affect the setting of
3747 compiler options, which is what is needed, but I have no means of testing
3748 whether it really works. (The user who reported this had patched the
3749 generated configure script, which of course I cannot do.)
3751 LATER: After change 22 below (new libtool files), the configure script
3752 seems to know about icc (and also ecc). Therefore, I have commented out
3753 this hack in configure.in.
3755 7. Added support for pkg-config (2 patches were sent in).
3757 8. Negated POSIX character classes that used a combination of internal tables
3758 were completely broken. These were [[:^alpha:]], [[:^alnum:]], and
3759 [[:^ascii]]. Typically, they would match almost any characters. The other
3760 POSIX classes were not broken in this way.
3762 9. Matching the pattern "\b.*?" against "ab cd", starting at offset 1, failed
3763 to find the match, as PCRE was deluded into thinking that the match had to
3764 start at the start point or following a newline. The same bug applied to
3765 patterns with negative forward assertions or any backward assertions
3766 preceding ".*" at the start, unless the pattern required a fixed first
3767 character. This was a failing pattern: "(?!.bcd).*". The bug is now fixed.
3769 10. In UTF-8 mode, when moving forwards in the subject after a failed match
3770 starting at the last subject character, bytes beyond the end of the subject
3771 string were read.
3773 11. Renamed the variable "class" as "classbits" to make life easier for C++
3774 users. (Previously there was a macro definition, but it apparently wasn't
3775 enough.)
3777 12. Added the new field "tables" to the extra data so that tables can be passed
3778 in at exec time, or the internal tables can be re-selected. This allows
3779 a compiled regex to be saved and re-used at a later time by a different
3780 program that might have everything at different addresses.
3782 13. Modified the pcre-config script so that, when run on Solaris, it shows a
3783 -R library as well as a -L library.
3785 14. The debugging options of pcretest (-d on the command line or D on a
3786 pattern) showed incorrect output for anything following an extended class
3787 that contained multibyte characters and which was followed by a quantifier.
3789 15. Added optional support for general category Unicode character properties
3790 via the \p, \P, and \X escapes. Unicode property support implies UTF-8
3791 support. It adds about 90K to the size of the library. The meanings of the
3792 inbuilt class escapes such as \d and \s have NOT been changed.
3794 16. Updated pcredemo.c to include calls to free() to release the memory for the
3795 compiled pattern.
3797 17. The generated file chartables.c was being created in the source directory
3798 instead of in the building directory. This caused the build to fail if the
3799 source directory was different from the building directory, and was
3800 read-only.
3802 18. Added some sample Win commands from Mark Tetrode into the NON-UNIX-USE
3803 file. No doubt somebody will tell me if they don't make sense... Also added
3804 Dan Mooney's comments about building on OpenVMS.
3806 19. Added support for partial matching via the PCRE_PARTIAL option for
3807 pcre_exec() and the \P data escape in pcretest.
3809 20. Extended pcretest with 3 new pattern features:
3811 (i) A pattern option of the form ">rest-of-line" causes pcretest to
3812 write the compiled pattern to the file whose name is "rest-of-line".
3813 This is a straight binary dump of the data, with the saved pointer to
3814 the character tables forced to be NULL. The study data, if any, is
3815 written too. After writing, pcretest reads a new pattern.
3817 (ii) If, instead of a pattern, "<rest-of-line" is given, pcretest reads a
3818 compiled pattern from the given file. There must not be any
3819 occurrences of "<" in the file name (pretty unlikely); if there are,
3820 pcretest will instead treat the initial "<" as a pattern delimiter.
3821 After reading in the pattern, pcretest goes on to read data lines as
3822 usual.
3824 (iii) The F pattern option causes pcretest to flip the bytes in the 32-bit
3825 and 16-bit fields in a compiled pattern, to simulate a pattern that
3826 was compiled on a host of opposite endianness.
3828 21. The pcre-exec() function can now cope with patterns that were compiled on
3829 hosts of opposite endianness, with this restriction:
3831 As for any compiled expression that is saved and used later, the tables
3832 pointer field cannot be preserved; the extra_data field in the arguments
3833 to pcre_exec() should be used to pass in a tables address if a value
3834 other than the default internal tables were used at compile time.
3836 22. Calling pcre_exec() with a negative value of the "ovecsize" parameter is
3837 now diagnosed as an error. Previously, most of the time, a negative number
3838 would have been treated as zero, but if in addition "ovector" was passed as
3839 NULL, a crash could occur.
3841 23. Updated the files ltmain.sh, config.sub, config.guess, and aclocal.m4 with
3842 new versions from the libtool 1.5 distribution (the last one is a copy of
3843 a file called libtool.m4). This seems to have fixed the need to patch
3844 "configure" to support Darwin 1.3 (which I used to do). However, I still
3845 had to patch ltmain.sh to ensure that ${SED} is set (it isn't on my
3846 workstation).
3848 24. Changed the PCRE licence to be the more standard "BSD" licence.
3851 Version 4.5 01-Dec-03
3852 ---------------------
3854 1. There has been some re-arrangement of the code for the match() function so
3855 that it can be compiled in a version that does not call itself recursively.
3856 Instead, it keeps those local variables that need separate instances for
3857 each "recursion" in a frame on the heap, and gets/frees frames whenever it
3858 needs to "recurse". Keeping track of where control must go is done by means
3859 of setjmp/longjmp. The whole thing is implemented by a set of macros that
3860 hide most of the details from the main code, and operates only if
3861 NO_RECURSE is defined while compiling pcre.c. If PCRE is built using the
3862 "configure" mechanism, "--disable-stack-for-recursion" turns on this way of
3863 operating.
3865 To make it easier for callers to provide specially tailored get/free
3866 functions for this usage, two new functions, pcre_stack_malloc, and
3867 pcre_stack_free, are used. They are always called in strict stacking order,
3868 and the size of block requested is always the same.
3870 The PCRE_CONFIG_STACKRECURSE info parameter can be used to find out whether
3871 PCRE has been compiled to use the stack or the heap for recursion. The
3872 -C option of pcretest uses this to show which version is compiled.
3874 A new data escape \S, is added to pcretest; it causes the amounts of store
3875 obtained and freed by both kinds of malloc/free at match time to be added
3876 to the output.
3878 2. Changed the locale test to use "fr_FR" instead of "fr" because that's
3879 what's available on my current Linux desktop machine.
3881 3. When matching a UTF-8 string, the test for a valid string at the start has
3882 been extended. If start_offset is not zero, PCRE now checks that it points
3883 to a byte that is the start of a UTF-8 character. If not, it returns
3884 PCRE_ERROR_BADUTF8_OFFSET (-11). Note: the whole string is still checked;
3885 this is necessary because there may be backward assertions in the pattern.
3886 When matching the same subject several times, it may save resources to use
3887 PCRE_NO_UTF8_CHECK on all but the first call if the string is long.
3889 4. The code for checking the validity of UTF-8 strings has been tightened so
3890 that it rejects (a) strings containing 0xfe or 0xff bytes and (b) strings
3891 containing "overlong sequences".
3893 5. Fixed a bug (appearing twice) that I could not find any way of exploiting!
3894 I had written "if ((digitab[*p++] && chtab_digit) == 0)" where the "&&"
3895 should have been "&", but it just so happened that all the cases this let
3896 through by mistake were picked up later in the function.
3898 6. I had used a variable called "isblank" - this is a C99 function, causing
3899 some compilers to warn. To avoid this, I renamed it (as "blankclass").
3901 7. Cosmetic: (a) only output another newline at the end of pcretest if it is
3902 prompting; (b) run "./pcretest /dev/null" at the start of the test script
3903 so the version is shown; (c) stop "make test" echoing "./RunTest".
3905 8. Added patches from David Burgess to enable PCRE to run on EBCDIC systems.
3907 9. The prototype for memmove() for systems that don't have it was using
3908 size_t, but the inclusion of the header that defines size_t was later. I've
3909 moved the #includes for the C headers earlier to avoid this.
3911 10. Added some adjustments to the code to make it easier to compiler on certain
3912 special systems:
3914 (a) Some "const" qualifiers were missing.
3915 (b) Added the macro EXPORT before all exported functions; by default this
3916 is defined to be empty.
3917 (c) Changed the dftables auxiliary program (that builds chartables.c) so
3918 that it reads its output file name as an argument instead of writing
3919 to the standard output and assuming this can be redirected.
3921 11. In UTF-8 mode, if a recursive reference (e.g. (?1)) followed a character
3922 class containing characters with values greater than 255, PCRE compilation
3923 went into a loop.
3925 12. A recursive reference to a subpattern that was within another subpattern
3926 that had a minimum quantifier of zero caused PCRE to crash. For example,
3927 (x(y(?2))z)? provoked this bug with a subject that got as far as the
3928 recursion. If the recursively-called subpattern itself had a zero repeat,
3929 that was OK.
3931 13. In pcretest, the buffer for reading a data line was set at 30K, but the
3932 buffer into which it was copied (for escape processing) was still set at
3933 1024, so long lines caused crashes.
3935 14. A pattern such as /[ab]{1,3}+/ failed to compile, giving the error
3936 "internal error: code overflow...". This applied to any character class
3937 that was followed by a possessive quantifier.
3939 15. Modified the Makefile to add libpcre.la as a prerequisite for
3940 libpcreposix.la because I was told this is needed for a parallel build to
3941 work.
3943 16. If a pattern that contained .* following optional items at the start was
3944 studied, the wrong optimizing data was generated, leading to matching
3945 errors. For example, studying /[ab]*.*c/ concluded, erroneously, that any
3946 matching string must start with a or b or c. The correct conclusion for
3947 this pattern is that a match can start with any character.
3950 Version 4.4 13-Aug-03
3951 ---------------------
3953 1. In UTF-8 mode, a character class containing characters with values between
3954 127 and 255 was not handled correctly if the compiled pattern was studied.
3955 In fixing this, I have also improved the studying algorithm for such
3956 classes (slightly).
3958 2. Three internal functions had redundant arguments passed to them. Removal
3959 might give a very teeny performance improvement.
3961 3. Documentation bug: the value of the capture_top field in a callout is *one
3962 more than* the number of the hightest numbered captured substring.
3964 4. The Makefile linked pcretest and pcregrep with -lpcre, which could result
3965 in incorrectly linking with a previously installed version. They now link
3966 explicitly with libpcre.la.
3968 5. configure.in no longer needs to recognize Cygwin specially.
3970 6. A problem in pcre.in for Windows platforms is fixed.
3972 7. If a pattern was successfully studied, and the -d (or /D) flag was given to
3973 pcretest, it used to include the size of the study block as part of its
3974 output. Unfortunately, the structure contains a field that has a different
3975 size on different hardware architectures. This meant that the tests that
3976 showed this size failed. As the block is currently always of a fixed size,
3977 this information isn't actually particularly useful in pcretest output, so
3978 I have just removed it.
3980 8. Three pre-processor statements accidentally did not start in column 1.
3981 Sadly, there are *still* compilers around that complain, even though
3982 standard C has not required this for well over a decade. Sigh.
3984 9. In pcretest, the code for checking callouts passed small integers in the
3985 callout_data field, which is a void * field. However, some picky compilers
3986 complained about the casts involved for this on 64-bit systems. Now
3987 pcretest passes the address of the small integer instead, which should get
3988 rid of the warnings.
3990 10. By default, when in UTF-8 mode, PCRE now checks for valid UTF-8 strings at
3991 both compile and run time, and gives an error if an invalid UTF-8 sequence
3992 is found. There is a option for disabling this check in cases where the
3993 string is known to be correct and/or the maximum performance is wanted.
3995 11. In response to a bug report, I changed one line in Makefile.in from
3997 -Wl,--out-implib,.libs/lib@WIN_PREFIX@pcreposix.dll.a \
3998 to
3999 -Wl,--out-implib,.libs/@WIN_PREFIX@libpcreposix.dll.a \
4001 to look similar to other lines, but I have no way of telling whether this
4002 is the right thing to do, as I do not use Windows. No doubt I'll get told
4003 if it's wrong...
4006 Version 4.3 21-May-03
4007 ---------------------
4009 1. Two instances of @WIN_PREFIX@ omitted from the Windows targets in the
4010 Makefile.
4012 2. Some refactoring to improve the quality of the code:
4014 (i) The utf8_table... variables are now declared "const".
4016 (ii) The code for \cx, which used the "case flipping" table to upper case
4017 lower case letters, now just substracts 32. This is ASCII-specific,
4018 but the whole concept of \cx is ASCII-specific, so it seems
4019 reasonable.
4021 (iii) PCRE was using its character types table to recognize decimal and
4022 hexadecimal digits in the pattern. This is silly, because it handles
4023 only 0-9, a-f, and A-F, but the character types table is locale-
4024 specific, which means strange things might happen. A private
4025 table is now used for this - though it costs 256 bytes, a table is
4026 much faster than multiple explicit tests. Of course, the standard
4027 character types table is still used for matching digits in subject
4028 strings against \d.
4030 (iv) Strictly, the identifier ESC_t is reserved by POSIX (all identifiers
4031 ending in _t are). So I've renamed it as ESC_tee.
4033 3. The first argument for regexec() in the POSIX wrapper should have been
4034 defined as "const".
4036 4. Changed pcretest to use malloc() for its buffers so that they can be
4037 Electric Fenced for debugging.
4039 5. There were several places in the code where, in UTF-8 mode, PCRE would try
4040 to read one or more bytes before the start of the subject string. Often this
4041 had no effect on PCRE's behaviour, but in some circumstances it could
4042 provoke a segmentation fault.
4044 6. A lookbehind at the start of a pattern in UTF-8 mode could also cause PCRE
4045 to try to read one or more bytes before the start of the subject string.
4047 7. A lookbehind in a pattern matched in non-UTF-8 mode on a PCRE compiled with
4048 UTF-8 support could misbehave in various ways if the subject string
4049 contained bytes with the 0x80 bit set and the 0x40 bit unset in a lookbehind
4050 area. (PCRE was not checking for the UTF-8 mode flag, and trying to move
4051 back over UTF-8 characters.)
4054 Version 4.2 14-Apr-03
4055 ---------------------
4057 1. Typo "#if SUPPORT_UTF8" instead of "#ifdef SUPPORT_UTF8" fixed.
4059 2. Changes to the building process, supplied by Ronald Landheer-Cieslak
4060 [ON_WINDOWS]: new variable, "#" on non-Windows platforms
4061 [NOT_ON_WINDOWS]: new variable, "#" on Windows platforms
4062 [WIN_PREFIX]: new variable, "cyg" for Cygwin
4063 * Makefile.in: use autoconf substitution for OBJEXT, EXEEXT, BUILD_OBJEXT
4065 Note: automatic setting of the BUILD variables is not yet working
4066 set CPPFLAGS and BUILD_CPPFLAGS (but don't use yet) - should be used at
4067 compile-time but not at link-time
4068 [LINK]: use for linking executables only
4069 make different versions for Windows and non-Windows
4070 [LINKLIB]: new variable, copy of UNIX-style LINK, used for linking
4071 libraries
4072 [LINK_FOR_BUILD]: new variable
4073 [OBJEXT]: use throughout
4074 [EXEEXT]: use throughout
4075 <winshared>: new target
4076 <wininstall>: new target
4077 <dftables.o>: use native compiler
4078 <dftables>: use native linker
4079 <install>: handle Windows platform correctly
4080 <clean>: ditto
4081 <check>: ditto
4082 copy DLL to top builddir before testing
4084 As part of these changes, -no-undefined was removed again. This was reported
4085 to give trouble on HP-UX 11.0, so getting rid of it seems like a good idea
4086 in any case.
4088 3. Some tidies to get rid of compiler warnings:
4090 . In the match_data structure, match_limit was an unsigned long int, whereas
4091 match_call_count was an int. I've made them both unsigned long ints.
4093 . In pcretest the fact that a const uschar * doesn't automatically cast to
4094 a void * provoked a warning.
4096 . Turning on some more compiler warnings threw up some "shadow" variables
4097 and a few more missing casts.
4099 4. If PCRE was complied with UTF-8 support, but called without the PCRE_UTF8
4100 option, a class that contained a single character with a value between 128
4101 and 255 (e.g. /[\xFF]/) caused PCRE to crash.
4103 5. If PCRE was compiled with UTF-8 support, but called without the PCRE_UTF8
4104 option, a class that contained several characters, but with at least one
4105 whose value was between 128 and 255 caused PCRE to crash.
4108 Version 4.1 12-Mar-03
4109 ---------------------
4111 1. Compiling with gcc -pedantic found a couple of places where casts were
4112 needed, and a string in dftables.c that was longer than standard compilers are
4113 required to support.
4115 2. Compiling with Sun's compiler found a few more places where the code could
4116 be tidied up in order to avoid warnings.
4118 3. The variables for cross-compiling were called HOST_CC and HOST_CFLAGS; the
4119 first of these names is deprecated in the latest Autoconf in favour of the name
4120 CC_FOR_BUILD, because "host" is typically used to mean the system on which the
4121 compiled code will be run. I can't find a reference for HOST_CFLAGS, but by
4122 analogy I have changed it to CFLAGS_FOR_BUILD.
4124 4. Added -no-undefined to the linking command in the Makefile, because this is
4125 apparently helpful for Windows. To make it work, also added "-L. -lpcre" to the
4126 linking step for the pcreposix library.
4128 5. PCRE was failing to diagnose the case of two named groups with the same
4129 name.
4131 6. A problem with one of PCRE's optimizations was discovered. PCRE remembers a
4132 literal character that is needed in the subject for a match, and scans along to
4133 ensure that it is present before embarking on the full matching process. This
4134 saves time in cases of nested unlimited repeats that are never going to match.
4135 Problem: the scan can take a lot of time if the subject is very long (e.g.
4136 megabytes), thus penalizing straightforward matches. It is now done only if the
4137 amount of subject to be scanned is less than 1000 bytes.
4139 7. A lesser problem with the same optimization is that it was recording the
4140 first character of an anchored pattern as "needed", thus provoking a search
4141 right along the subject, even when the first match of the pattern was going to
4142 fail. The "needed" character is now not set for anchored patterns, unless it
4143 follows something in the pattern that is of non-fixed length. Thus, it still
4144 fulfils its original purpose of finding quick non-matches in cases of nested
4145 unlimited repeats, but isn't used for simple anchored patterns such as /^abc/.
4148 Version 4.0 17-Feb-03
4149 ---------------------
4151 1. If a comment in an extended regex that started immediately after a meta-item
4152 extended to the end of string, PCRE compiled incorrect data. This could lead to
4153 all kinds of weird effects. Example: /#/ was bad; /()#/ was bad; /a#/ was not.
4155 2. Moved to autoconf 2.53 and libtool 1.4.2.
4157 3. Perl 5.8 no longer needs "use utf8" for doing UTF-8 things. Consequently,
4158 the special perltest8 script is no longer needed - all the tests can be run
4159 from a single perltest script.
4161 4. From 5.004, Perl has not included the VT character (0x0b) in the set defined
4162 by \s. It has now been removed in PCRE. This means it isn't recognized as
4163 whitespace in /x regexes too, which is the same as Perl. Note that the POSIX
4164 class [:space:] *does* include VT, thereby creating a mess.
4166 5. Added the class [:blank:] (a GNU extension from Perl 5.8) to match only
4167 space and tab.
4169 6. Perl 5.005 was a long time ago. It's time to amalgamate the tests that use
4170 its new features into the main test script, reducing the number of scripts.
4172 7. Perl 5.8 has changed the meaning of patterns like /a(?i)b/. Earlier versions
4173 were backward compatible, and made the (?i) apply to the whole pattern, as if
4174 /i were given. Now it behaves more logically, and applies the option setting
4175 only to what follows. PCRE has been changed to follow suit. However, if it
4176 finds options settings right at the start of the pattern, it extracts them into
4177 the global options, as before. Thus, they show up in the info data.
4179 8. Added support for the \Q...\E escape sequence. Characters in between are
4180 treated as literals. This is slightly different from Perl in that $ and @ are
4181 also handled as literals inside the quotes. In Perl, they will cause variable
4182 interpolation. Note the following examples:
4184 Pattern PCRE matches Perl matches
4186 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
4187 \Qabc\$xyz\E abc\$xyz abc\$xyz
4188 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
4190 For compatibility with Perl, \Q...\E sequences are recognized inside character
4191 classes as well as outside them.
4193 9. Re-organized 3 code statements in pcretest to avoid "overflow in
4194 floating-point constant arithmetic" warnings from a Microsoft compiler. Added a
4195 (size_t) cast to one statement in pcretest and one in pcreposix to avoid
4196 signed/unsigned warnings.
4198 10. SunOS4 doesn't have strtoul(). This was used only for unpicking the -o
4199 option for pcretest, so I've replaced it by a simple function that does just
4200 that job.
4202 11. pcregrep was ending with code 0 instead of 2 for the commands "pcregrep" or
4203 "pcregrep -".
4205 12. Added "possessive quantifiers" ?+, *+, ++, and {,}+ which come from Sun's
4206 Java package. This provides some syntactic sugar for simple cases of what my
4207 documentation calls "once-only subpatterns". A pattern such as x*+ is the same
4208 as (?>x*). In other words, if what is inside (?>...) is just a single repeated
4209 item, you can use this simplified notation. Note that only makes sense with
4210 greedy quantifiers. Consequently, the use of the possessive quantifier forces
4211 greediness, whatever the setting of the PCRE_UNGREEDY option.
4213 13. A change of greediness default within a pattern was not taking effect at
4214 the current level for patterns like /(b+(?U)a+)/. It did apply to parenthesized
4215 subpatterns that followed. Patterns like /b+(?U)a+/ worked because the option
4216 was abstracted outside.
4218 14. PCRE now supports the \G assertion. It is true when the current matching
4219 position is at the start point of the match. This differs from \A when the
4220 starting offset is non-zero. Used with the /g option of pcretest (or similar
4221 code), it works in the same way as it does for Perl's /g option. If all
4222 alternatives of a regex begin with \G, the expression is anchored to the start
4223 match position, and the "anchored" flag is set in the compiled expression.
4225 15. Some bugs concerning the handling of certain option changes within patterns
4226 have been fixed. These applied to options other than (?ims). For example,
4227 "a(?x: b c )d" did not match "XabcdY" but did match "Xa b c dY". It should have
4228 been the other way round. Some of this was related to change 7 above.
4230 16. PCRE now gives errors for /[.x.]/ and /[=x=]/ as unsupported POSIX
4231 features, as Perl does. Previously, PCRE gave the warnings only for /[[.x.]]/
4232 and /[[=x=]]/. PCRE now also gives an error for /[:name:]/ because it supports
4233 POSIX classes only within a class (e.g. /[[:alpha:]]/).
4235 17. Added support for Perl's \C escape. This matches one byte, even in UTF8
4236 mode. Unlike ".", it always matches newline, whatever the setting of
4237 PCRE_DOTALL. However, PCRE does not permit \C to appear in lookbehind
4238 assertions. Perl allows it, but it doesn't (in general) work because it can't
4239 calculate the length of the lookbehind. At least, that's the case for Perl
4240 5.8.0 - I've been told they are going to document that it doesn't work in
4241 future.
4243 18. Added an error diagnosis for escapes that PCRE does not support: these are
4244 \L, \l, \N, \P, \p, \U, \u, and \X.
4246 19. Although correctly diagnosing a missing ']' in a character class, PCRE was
4247 reading past the end of the pattern in cases such as /[abcd/.
4249 20. PCRE was getting more memory than necessary for patterns with classes that
4250 contained both POSIX named classes and other characters, e.g. /[[:space:]abc/.
4252 21. Added some code, conditional on #ifdef VPCOMPAT, to make life easier for
4253 compiling PCRE for use with Virtual Pascal.
4255 22. Small fix to the Makefile to make it work properly if the build is done
4256 outside the source tree.
4258 23. Added a new extension: a condition to go with recursion. If a conditional
4259 subpattern starts with (?(R) the "true" branch is used if recursion has
4260 happened, whereas the "false" branch is used only at the top level.
4262 24. When there was a very long string of literal characters (over 255 bytes
4263 without UTF support, over 250 bytes with UTF support), the computation of how
4264 much memory was required could be incorrect, leading to segfaults or other
4265 strange effects.
4267 25. PCRE was incorrectly assuming anchoring (either to start of subject or to
4268 start of line for a non-DOTALL pattern) when a pattern started with (.*) and
4269 there was a subsequent back reference to those brackets. This meant that, for
4270 example, /(.*)\d+\1/ failed to match "abc123bc". Unfortunately, it isn't
4271 possible to check for precisely this case. All we can do is abandon the
4272 optimization if .* occurs inside capturing brackets when there are any back
4273 references whatsoever. (See below for a better fix that came later.)
4275 26. The handling of the optimization for finding the first character of a
4276 non-anchored pattern, and for finding a character that is required later in the
4277 match were failing in some cases. This didn't break the matching; it just
4278 failed to optimize when it could. The way this is done has been re-implemented.
4280 27. Fixed typo in error message for invalid (?R item (it said "(?p").
4282 28. Added a new feature that provides some of the functionality that Perl
4283 provides with (?{...}). The facility is termed a "callout". The way it is done
4284 in PCRE is for the caller to provide an optional function, by setting
4285 pcre_callout to its entry point. Like pcre_malloc and pcre_free, this is a
4286 global variable. By default it is unset, which disables all calling out. To get
4287 the function called, the regex must include (?C) at appropriate points. This
4288 is, in fact, equivalent to (?C0), and any number <= 255 may be given with (?C).
4289 This provides a means of identifying different callout points. When PCRE
4290 reaches such a point in the regex, if pcre_callout has been set, the external
4291 function is called. It is provided with data in a structure called
4292 pcre_callout_block, which is defined in pcre.h. If the function returns 0,
4293 matching continues; if it returns a non-zero value, the match at the current
4294 point fails. However, backtracking will occur if possible. [This was changed
4295 later and other features added - see item 49 below.]
4297 29. pcretest is upgraded to test the callout functionality. It provides a
4298 callout function that displays information. By default, it shows the start of
4299 the match and the current position in the text. There are some new data escapes
4300 to vary what happens:
4302 \C+ in addition, show current contents of captured substrings
4303 \C- do not supply a callout function
4304 \C!n return 1 when callout number n is reached
4305 \C!n!m return 1 when callout number n is reached for the mth time
4307 30. If pcregrep was called with the -l option and just a single file name, it
4308 output "<stdin>" if a match was found, instead of the file name.
4310 31. Improve the efficiency of the POSIX API to PCRE. If the number of capturing
4311 slots is less than POSIX_MALLOC_THRESHOLD, use a block on the stack to pass to
4312 pcre_exec(). This saves a malloc/free per call. The default value of
4313 POSIX_MALLOC_THRESHOLD is 10; it can be changed by --with-posix-malloc-threshold
4314 when configuring.
4316 32. The default maximum size of a compiled pattern is 64K. There have been a
4317 few cases of people hitting this limit. The code now uses macros to handle the
4318 storing of links as offsets within the compiled pattern. It defaults to 2-byte
4319 links, but this can be changed to 3 or 4 bytes by --with-link-size when
4320 configuring. Tests 2 and 5 work only with 2-byte links because they output
4321 debugging information about compiled patterns.
4323 33. Internal code re-arrangements:
4325 (a) Moved the debugging function for printing out a compiled regex into
4326 its own source file (printint.c) and used #include to pull it into
4327 pcretest.c and, when DEBUG is defined, into pcre.c, instead of having two
4328 separate copies.
4330 (b) Defined the list of op-code names for debugging as a macro in
4331 internal.h so that it is next to the definition of the opcodes.
4333 (c) Defined a table of op-code lengths for simpler skipping along compiled
4334 code. This is again a macro in internal.h so that it is next to the
4335 definition of the opcodes.
4337 34. Added support for recursive calls to individual subpatterns, along the
4338 lines of Robin Houston's patch (but implemented somewhat differently).
4340 35. Further mods to the Makefile to help Win32. Also, added code to pcregrep to
4341 allow it to read and process whole directories in Win32. This code was
4342 contributed by Lionel Fourquaux; it has not been tested by me.
4344 36. Added support for named subpatterns. The Python syntax (?P<name>...) is
4345 used to name a group. Names consist of alphanumerics and underscores, and must
4346 be unique. Back references use the syntax (?P=name) and recursive calls use
4347 (?P>name) which is a PCRE extension to the Python extension. Groups still have
4348 numbers. The function pcre_fullinfo() can be used after compilation to extract
4349 a name/number map. There are three relevant calls:
4351 PCRE_INFO_NAMEENTRYSIZE yields the size of each entry in the map
4352 PCRE_INFO_NAMECOUNT yields the number of entries
4353 PCRE_INFO_NAMETABLE yields a pointer to the map.
4355 The map is a vector of fixed-size entries. The size of each entry depends on
4356 the length of the longest name used. The first two bytes of each entry are the
4357 group number, most significant byte first. There follows the corresponding
4358 name, zero terminated. The names are in alphabetical order.
4360 37. Make the maximum literal string in the compiled code 250 for the non-UTF-8
4361 case instead of 255. Making it the same both with and without UTF-8 support
4362 means that the same test output works with both.
4364 38. There was a case of malloc(0) in the POSIX testing code in pcretest. Avoid
4365 calling malloc() with a zero argument.
4367 39. Change 25 above had to resort to a heavy-handed test for the .* anchoring
4368 optimization. I've improved things by keeping a bitmap of backreferences with
4369 numbers 1-31 so that if .* occurs inside capturing brackets that are not in
4370 fact referenced, the optimization can be applied. It is unlikely that a
4371 relevant occurrence of .* (i.e. one which might indicate anchoring or forcing
4372 the match to follow \n) will appear inside brackets with a number greater than
4373 31, but if it does, any back reference > 31 suppresses the optimization.
4375 40. Added a new compile-time option PCRE_NO_AUTO_CAPTURE. This has the effect
4376 of disabling numbered capturing parentheses. Any opening parenthesis that is
4377 not followed by ? behaves as if it were followed by ?: but named parentheses
4378 can still be used for capturing (and they will acquire numbers in the usual
4379 way).
4381 41. Redesigned the return codes from the match() function into yes/no/error so
4382 that errors can be passed back from deep inside the nested calls. A malloc
4383 failure while inside a recursive subpattern call now causes the
4384 PCRE_ERROR_NOMEMORY return instead of quietly going wrong.
4386 42. It is now possible to set a limit on the number of times the match()
4387 function is called in a call to pcre_exec(). This facility makes it possible to
4388 limit the amount of recursion and backtracking, though not in a directly
4389 obvious way, because the match() function is used in a number of different
4390 circumstances. The count starts from zero for each position in the subject
4391 string (for non-anchored patterns). The default limit is, for compatibility, a
4392 large number, namely 10 000 000. You can change this in two ways:
4394 (a) When configuring PCRE before making, you can use --with-match-limit=n
4395 to set a default value for the compiled library.
4397 (b) For each call to pcre_exec(), you can pass a pcre_extra block in which
4398 a different value is set. See 45 below.
4400 If the limit is exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
4402 43. Added a new function pcre_config(int, void *) to enable run-time extraction
4403 of things that can be changed at compile time. The first argument specifies
4404 what is wanted and the second points to where the information is to be placed.
4405 The current list of available information is:
4409 The output is an integer that is set to one if UTF-8 support is available;
4410 otherwise it is set to zero.
4414 The output is an integer that it set to the value of the code that is used for
4415 newline. It is either LF (10) or CR (13).
4419 The output is an integer that contains the number of bytes used for internal
4420 linkage in compiled expressions. The value is 2, 3, or 4. See item 32 above.
4424 The output is an integer that contains the threshold above which the POSIX
4425 interface uses malloc() for output vectors. See item 31 above.
4429 The output is an unsigned integer that contains the default limit of the number
4430 of match() calls in a pcre_exec() execution. See 42 above.
4432 44. pcretest has been upgraded by the addition of the -C option. This causes it
4433 to extract all the available output from the new pcre_config() function, and to
4434 output it. The program then exits immediately.
4436 45. A need has arisen to pass over additional data with calls to pcre_exec() in
4437 order to support additional features. One way would have been to define
4438 pcre_exec2() (for example) with extra arguments, but this would not have been
4439 extensible, and would also have required all calls to the original function to
4440 be mapped to the new one. Instead, I have chosen to extend the mechanism that
4441 is used for passing in "extra" data from pcre_study().
4443 The pcre_extra structure is now exposed and defined in pcre.h. It currently
4444 contains the following fields:
4446 flags a bitmap indicating which of the following fields are set
4447 study_data opaque data from pcre_study()
4448 match_limit a way of specifying a limit on match() calls for a specific
4449 call to pcre_exec()
4450 callout_data data for callouts (see 49 below)
4452 The flag bits are also defined in pcre.h, and are
4458 The pcre_study() function now returns one of these new pcre_extra blocks, with
4459 the actual study data pointed to by the study_data field, and the
4460 PCRE_EXTRA_STUDY_DATA flag set. This can be passed directly to pcre_exec() as
4461 before. That is, this change is entirely upwards-compatible and requires no
4462 change to existing code.
4464 If you want to pass in additional data to pcre_exec(), you can either place it
4465 in a pcre_extra block provided by pcre_study(), or create your own pcre_extra
4466 block.
4468 46. pcretest has been extended to test the PCRE_EXTRA_MATCH_LIMIT feature. If a
4469 data string contains the escape sequence \M, pcretest calls pcre_exec() several
4470 times with different match limits, until it finds the minimum value needed for
4471 pcre_exec() to complete. The value is then output. This can be instructive; for
4472 most simple matches the number is quite small, but for pathological cases it
4473 gets very large very quickly.
4475 47. There's a new option for pcre_fullinfo() called PCRE_INFO_STUDYSIZE. It
4476 returns the size of the data block pointed to by the study_data field in a
4477 pcre_extra block, that is, the value that was passed as the argument to
4478 pcre_malloc() when PCRE was getting memory in which to place the information
4479 created by pcre_study(). The fourth argument should point to a size_t variable.
4480 pcretest has been extended so that this information is shown after a successful
4481 pcre_study() call when information about the compiled regex is being displayed.
4483 48. Cosmetic change to Makefile: there's no need to have / after $(DESTDIR)
4484 because what follows is always an absolute path. (Later: it turns out that this
4485 is more than cosmetic for MinGW, because it doesn't like empty path
4486 components.)
4488 49. Some changes have been made to the callout feature (see 28 above):
4490 (i) A callout function now has three choices for what it returns:
4492 0 => success, carry on matching
4493 > 0 => failure at this point, but backtrack if possible
4494 < 0 => serious error, return this value from pcre_exec()
4496 Negative values should normally be chosen from the set of PCRE_ERROR_xxx
4497 values. In particular, returning PCRE_ERROR_NOMATCH forces a standard
4498 "match failed" error. The error number PCRE_ERROR_CALLOUT is reserved for
4499 use by callout functions. It will never be used by PCRE itself.
4501 (ii) The pcre_extra structure (see 45 above) has a void * field called
4502 callout_data, with corresponding flag bit PCRE_EXTRA_CALLOUT_DATA. The
4503 pcre_callout_block structure has a field of the same name. The contents of
4504 the field passed in the pcre_extra structure are passed to the callout
4505 function in the corresponding field in the callout block. This makes it
4506 easier to use the same callout-containing regex from multiple threads. For
4507 testing, the pcretest program has a new data escape
4509 \C*n pass the number n (may be negative) as callout_data
4511 If the callout function in pcretest receives a non-zero value as
4512 callout_data, it returns that value.
4514 50. Makefile wasn't handling CFLAGS properly when compiling dftables. Also,
4515 there were some redundant $(CFLAGS) in commands that are now specified as
4516 $(LINK), which already includes $(CFLAGS).
4518 51. Extensions to UTF-8 support are listed below. These all apply when (a) PCRE
4519 has been compiled with UTF-8 support *and* pcre_compile() has been compiled
4520 with the PCRE_UTF8 flag. Patterns that are compiled without that flag assume
4521 one-byte characters throughout. Note that case-insensitive matching applies
4522 only to characters whose values are less than 256. PCRE doesn't support the
4523 notion of cases for higher-valued characters.
4525 (i) A character class whose characters are all within 0-255 is handled as
4526 a bit map, and the map is inverted for negative classes. Previously, a
4527 character > 255 always failed to match such a class; however it should
4528 match if the class was a negative one (e.g. [^ab]). This has been fixed.
4530 (ii) A negated character class with a single character < 255 is coded as
4531 "not this character" (OP_NOT). This wasn't working properly when the test
4532 character was multibyte, either singly or repeated.
4534 (iii) Repeats of multibyte characters are now handled correctly in UTF-8
4535 mode, for example: \x{100}{2,3}.
4537 (iv) The character escapes \b, \B, \d, \D, \s, \S, \w, and \W (either
4538 singly or repeated) now correctly test multibyte characters. However,
4539 PCRE doesn't recognize any characters with values greater than 255 as
4540 digits, spaces, or word characters. Such characters always match \D, \S,
4541 and \W, and never match \d, \s, or \w.
4543 (v) Classes may now contain characters and character ranges with values
4544 greater than 255. For example: [ab\x{100}-\x{400}].
4546 (vi) pcregrep now has a --utf-8 option (synonym -u) which makes it call
4547 PCRE in UTF-8 mode.
4549 52. The info request value PCRE_INFO_FIRSTCHAR has been renamed
4550 PCRE_INFO_FIRSTBYTE because it is a byte value. However, the old name is
4551 retained for backwards compatibility. (Note that LASTLITERAL is also a byte
4552 value.)
4554 53. The single man page has become too large. I have therefore split it up into
4555 a number of separate man pages. These also give rise to individual HTML pages;
4556 these are now put in a separate directory, and there is an index.html page that
4557 lists them all. Some hyperlinking between the pages has been installed.
4559 54. Added convenience functions for handling named capturing parentheses.
4561 55. Unknown escapes inside character classes (e.g. [\M]) and escapes that
4562 aren't interpreted therein (e.g. [\C]) are literals in Perl. This is now also
4563 true in PCRE, except when the PCRE_EXTENDED option is set, in which case they
4564 are faulted.
4566 56. Introduced HOST_CC and HOST_CFLAGS which can be set in the environment when
4567 calling configure. These values are used when compiling the dftables.c program
4568 which is run to generate the source of the default character tables. They
4569 default to the values of CC and CFLAGS. If you are cross-compiling PCRE,
4570 you will need to set these values.
4572 57. Updated the building process for Windows DLL, as provided by Fred Cox.
4575 Version 3.9 02-Jan-02
4576 ---------------------
4578 1. A bit of extraneous text had somehow crept into the pcregrep documentation.
4580 2. If --disable-static was given, the building process failed when trying to
4581 build pcretest and pcregrep. (For some reason it was using libtool to compile
4582 them, which is not right, as they aren't part of the library.)
4585 Version 3.8 18-Dec-01
4586 ---------------------
4588 1. The experimental UTF-8 code was completely screwed up. It was packing the
4589 bytes in the wrong order. How dumb can you get?
4592 Version 3.7 29-Oct-01
4593 ---------------------
4595 1. In updating pcretest to check change 1 of version 3.6, I screwed up.
4596 This caused pcretest, when used on the test data, to segfault. Unfortunately,
4597 this didn't happen under Solaris 8, where I normally test things.
4599 2. The Makefile had to be changed to make it work on BSD systems, where 'make'
4600 doesn't seem to recognize that ./xxx and xxx are the same file. (This entry
4601 isn't in ChangeLog distributed with 3.7 because I forgot when I hastily made
4602 this fix an hour or so after the initial 3.7 release.)
4605 Version 3.6 23-Oct-01
4606 ---------------------
4608 1. Crashed with /(sens|respons)e and \1ibility/ and "sense and sensibility" if
4609 offsets passed as NULL with zero offset count.
4611 2. The config.guess and config.sub files had not been updated when I moved to
4612 the latest autoconf.
4615 Version 3.5 15-Aug-01
4616 ---------------------
4618 1. Added some missing #if !defined NOPOSIX conditionals in pcretest.c that
4619 had been forgotten.
4621 2. By using declared but undefined structures, we can avoid using "void"
4622 definitions in pcre.h while keeping the internal definitions of the structures
4623 private.
4625 3. The distribution is now built using autoconf 2.50 and libtool 1.4. From a
4626 user point of view, this means that both static and shared libraries are built
4627 by default, but this can be individually controlled. More of the work of
4628 handling this static/shared cases is now inside libtool instead of PCRE's make
4629 file.
4631 4. The pcretest utility is now installed along with pcregrep because it is
4632 useful for users (to test regexs) and by doing this, it automatically gets
4633 relinked by libtool. The documentation has been turned into a man page, so
4634 there are now .1, .txt, and .html versions in /doc.
4636 5. Upgrades to pcregrep:
4637 (i) Added long-form option names like gnu grep.
4638 (ii) Added --help to list all options with an explanatory phrase.
4639 (iii) Added -r, --recursive to recurse into sub-directories.
4640 (iv) Added -f, --file to read patterns from a file.
4642 6. pcre_exec() was referring to its "code" argument before testing that
4643 argument for NULL (and giving an error if it was NULL).
4645 7. Upgraded Makefile.in to allow for compiling in a different directory from
4646 the source directory.
4648 8. Tiny buglet in pcretest: when pcre_fullinfo() was called to retrieve the
4649 options bits, the pointer it was passed was to an int instead of to an unsigned
4650 long int. This mattered only on 64-bit systems.
4652 9. Fixed typo (3.4/1) in pcre.h again. Sigh. I had changed pcre.h (which is
4653 generated) instead of pcre.in, which it its source. Also made the same change
4654 in several of the .c files.
4656 10. A new release of gcc defines printf() as a macro, which broke pcretest
4657 because it had an ifdef in the middle of a string argument for printf(). Fixed
4658 by using separate calls to printf().
4660 11. Added --enable-newline-is-cr and --enable-newline-is-lf to the configure
4661 script, to force use of CR or LF instead of \n in the source. On non-Unix
4662 systems, the value can be set in config.h.
4664 12. The limit of 200 on non-capturing parentheses is a _nesting_ limit, not an
4665 absolute limit. Changed the text of the error message to make this clear, and
4666 likewise updated the man page.
4668 13. The limit of 99 on the number of capturing subpatterns has been removed.
4669 The new limit is 65535, which I hope will not be a "real" limit.
4672 Version 3.4 22-Aug-00
4673 ---------------------
4675 1. Fixed typo in pcre.h: unsigned const char * changed to const unsigned char *.
4677 2. Diagnose condition (?(0) as an error instead of crashing on matching.
4680 Version 3.3 01-Aug-00
4681 ---------------------
4683 1. If an octal character was given, but the value was greater than \377, it
4684 was not getting masked to the least significant bits, as documented. This could
4685 lead to crashes in some systems.
4687 2. Perl 5.6 (if not earlier versions) accepts classes like [a-\d] and treats
4688 the hyphen as a literal. PCRE used to give an error; it now behaves like Perl.
4690 3. Added the functions pcre_free_substring() and pcre_free_substring_list().
4691 These just pass their arguments on to (pcre_free)(), but they are provided
4692 because some uses of PCRE bind it to non-C systems that can call its functions,
4693 but cannot call free() or pcre_free() directly.
4695 4. Add "make test" as a synonym for "make check". Corrected some comments in
4696 the Makefile.
4698 5. Add $(DESTDIR)/ in front of all the paths in the "install" target in the
4699 Makefile.
4701 6. Changed the name of pgrep to pcregrep, because Solaris has introduced a
4702 command called pgrep for grepping around the active processes.
4704 7. Added the beginnings of support for UTF-8 character strings.
4706 8. Arranged for the Makefile to pass over the settings of CC, CFLAGS, and
4707 RANLIB to ./ltconfig so that they are used by libtool. I think these are all
4708 the relevant ones. (AR is not passed because ./ltconfig does its own figuring
4709 out for the ar command.)
4712 Version 3.2 12-May-00
4713 ---------------------
4715 This is purely a bug fixing release.
4717 1. If the pattern /((Z)+|A)*/ was matched agained ZABCDEFG it matched Z instead
4718 of ZA. This was just one example of several cases that could provoke this bug,
4719 which was introduced by change 9 of version 2.00. The code for breaking
4720 infinite loops after an iteration that matches an empty string was't working
4721 correctly.
4723 2. The pcretest program was not imitating Perl correctly for the pattern /a*/g
4724 when matched against abbab (for example). After matching an empty string, it
4725 wasn't forcing anchoring when setting PCRE_NOTEMPTY for the next attempt; this
4726 caused it to match further down the string than it should.
4728 3. The code contained an inclusion of sys/types.h. It isn't clear why this
4729 was there because it doesn't seem to be needed, and it causes trouble on some
4730 systems, as it is not a Standard C header. It has been removed.
4732 4. Made 4 silly changes to the source to avoid stupid compiler warnings that
4733 were reported on the Macintosh. The changes were from
4735 while ((c = *(++ptr)) != 0 && c != '\n');
4736 to
4737 while ((c = *(++ptr)) != 0 && c != '\n') ;
4739 Totally extraordinary, but if that's what it takes...
4741 5. PCRE is being used in one environment where neither memmove() nor bcopy() is
4742 available. Added HAVE_BCOPY and an autoconf test for it; if neither
4743 HAVE_MEMMOVE nor HAVE_BCOPY is set, use a built-in emulation function which
4744 assumes the way PCRE uses memmove() (always moving upwards).
4746 6. PCRE is being used in one environment where strchr() is not available. There
4747 was only one use in pcre.c, and writing it out to avoid strchr() probably gives
4748 faster code anyway.
4751 Version 3.1 09-Feb-00
4752 ---------------------
4754 The only change in this release is the fixing of some bugs in Makefile.in for
4755 the "install" target:
4757 (1) It was failing to install pcreposix.h.
4759 (2) It was overwriting the pcre.3 man page with the pcreposix.3 man page.
4762 Version 3.0 01-Feb-00
4763 ---------------------
4765 1. Add support for the /+ modifier to perltest (to output $` like it does in
4766 pcretest).
4768 2. Add support for the /g modifier to perltest.
4770 3. Fix pcretest so that it behaves even more like Perl for /g when the pattern
4771 matches null strings.
4773 4. Fix perltest so that it doesn't do unwanted things when fed an empty
4774 pattern. Perl treats empty patterns specially - it reuses the most recent
4775 pattern, which is not what we want. Replace // by /(?#)/ in order to avoid this
4776 effect.
4778 5. The POSIX interface was broken in that it was just handing over the POSIX
4779 captured string vector to pcre_exec(), but (since release 2.00) PCRE has
4780 required a bigger vector, with some working space on the end. This means that
4781 the POSIX wrapper now has to get and free some memory, and copy the results.
4783 6. Added some simple autoconf support, placing the test data and the
4784 documentation in separate directories, re-organizing some of the
4785 information files, and making it build pcre-config (a GNU standard). Also added
4786 libtool support for building PCRE as a shared library, which is now the
4787 default.
4789 7. Got rid of the leading zero in the definition of PCRE_MINOR because 08 and
4790 09 are not valid octal constants. Single digits will be used for minor values
4791 less than 10.
4793 8. Defined REG_EXTENDED and REG_NOSUB as zero in the POSIX header, so that
4794 existing programs that set these in the POSIX interface can use PCRE without
4795 modification.
4797 9. Added a new function, pcre_fullinfo() with an extensible interface. It can
4798 return all that pcre_info() returns, plus additional data. The pcre_info()
4799 function is retained for compatibility, but is considered to be obsolete.
4801 10. Added experimental recursion feature (?R) to handle one common case that
4802 Perl 5.6 will be able to do with (?p{...}).
4804 11. Added support for POSIX character classes like [:alpha:], which Perl is
4805 adopting.
4808 Version 2.08 31-Aug-99
4809 ----------------------
4811 1. When startoffset was not zero and the pattern began with ".*", PCRE was not
4812 trying to match at the startoffset position, but instead was moving forward to
4813 the next newline as if a previous match had failed.
4815 2. pcretest was not making use of PCRE_NOTEMPTY when repeating for /g and /G,
4816 and could get into a loop if a null string was matched other than at the start
4817 of the subject.
4819 3. Added definitions of PCRE_MAJOR and PCRE_MINOR to pcre.h so the version can
4820 be distinguished at compile time, and for completeness also added PCRE_DATE.
4822 5. Added Paul Sokolovsky's minor changes to make it easy to compile a Win32 DLL
4823 in GnuWin32 environments.
4826 Version 2.07 29-Jul-99
4827 ----------------------
4829 1. The documentation is now supplied in plain text form and HTML as well as in
4830 the form of man page sources.
4832 2. C++ compilers don't like assigning (void *) values to other pointer types.
4833 In particular this affects malloc(). Although there is no problem in Standard
4834 C, I've put in casts to keep C++ compilers happy.
4836 3. Typo on pcretest.c; a cast of (unsigned char *) in the POSIX regexec() call
4837 should be (const char *).
4839 4. If NOPOSIX is defined, pcretest.c compiles without POSIX support. This may
4840 be useful for non-Unix systems who don't want to bother with the POSIX stuff.
4841 However, I haven't made this a standard facility. The documentation doesn't
4842 mention it, and the Makefile doesn't support it.
4844 5. The Makefile now contains an "install" target, with editable destinations at
4845 the top of the file. The pcretest program is not installed.
4847 6. pgrep -V now gives the PCRE version number and date.
4849 7. Fixed bug: a zero repetition after a literal string (e.g. /abcde{0}/) was
4850 causing the entire string to be ignored, instead of just the last character.
4852 8. If a pattern like /"([^\\"]+|\\.)*"/ is applied in the normal way to a
4853 non-matching string, it can take a very, very long time, even for strings of
4854 quite modest length, because of the nested recursion. PCRE now does better in
4855 some of these cases. It does this by remembering the last required literal
4856 character in the pattern, and pre-searching the subject to ensure it is present
4857 before running the real match. In other words, it applies a heuristic to detect
4858 some types of certain failure quickly, and in the above example, if presented
4859 with a string that has no trailing " it gives "no match" very quickly.
4861 9. A new runtime option PCRE_NOTEMPTY causes null string matches to be ignored;
4862 other alternatives are tried instead.
4865 Version 2.06 09-Jun-99
4866 ----------------------
4868 1. Change pcretest's output for amount of store used to show just the code
4869 space, because the remainder (the data block) varies in size between 32-bit and
4870 64-bit systems.
4872 2. Added an extra argument to pcre_exec() to supply an offset in the subject to
4873 start matching at. This allows lookbehinds to work when searching for multiple
4874 occurrences in a string.
4876 3. Added additional options to pcretest for testing multiple occurrences:
4878 /+ outputs the rest of the string that follows a match
4879 /g loops for multiple occurrences, using the new startoffset argument
4880 /G loops for multiple occurrences by passing an incremented pointer
4882 4. PCRE wasn't doing the "first character" optimization for patterns starting
4883 with \b or \B, though it was doing it for other lookbehind assertions. That is,
4884 it wasn't noticing that a match for a pattern such as /\bxyz/ has to start with
4885 the letter 'x'. On long subject strings, this gives a significant speed-up.
4888 Version 2.05 21-Apr-99
4889 ----------------------
4891 1. Changed the type of magic_number from int to long int so that it works
4892 properly on 16-bit systems.
4894 2. Fixed a bug which caused patterns starting with .* not to work correctly
4895 when the subject string contained newline characters. PCRE was assuming
4896 anchoring for such patterns in all cases, which is not correct because .* will
4897 not pass a newline unless PCRE_DOTALL is set. It now assumes anchoring only if
4898 DOTALL is set at top level; otherwise it knows that patterns starting with .*
4899 must be retried after every newline in the subject.
4902 Version 2.04 18-Feb-99
4903 ----------------------
4905 1. For parenthesized subpatterns with repeats whose minimum was zero, the
4906 computation of the store needed to hold the pattern was incorrect (too large).
4907 If such patterns were nested a few deep, this could multiply and become a real
4908 problem.
4910 2. Added /M option to pcretest to show the memory requirement of a specific
4911 pattern. Made -m a synonym of -s (which does this globally) for compatibility.
4913 3. Subpatterns of the form (regex){n,m} (i.e. limited maximum) were being
4914 compiled in such a way that the backtracking after subsequent failure was
4915 pessimal. Something like (a){0,3} was compiled as (a)?(a)?(a)? instead of
4916 ((a)((a)(a)?)?)? with disastrous performance if the maximum was of any size.
4919 Version 2.03 02-Feb-99
4920 ----------------------
4922 1. Fixed typo and small mistake in man page.
4924 2. Added 4th condition (GPL supersedes if conflict) and created separate
4925 LICENCE file containing the conditions.
4927 3. Updated pcretest so that patterns such as /abc\/def/ work like they do in
4928 Perl, that is the internal \ allows the delimiter to be included in the
4929 pattern. Locked out the use of \ as a delimiter. If \ immediately follows
4930 the final delimiter, add \ to the end of the pattern (to test the error).
4932 4. Added the convenience functions for extracting substrings after a successful
4933 match. Updated pcretest to make it able to test these functions.
4936 Version 2.02 14-Jan-99
4937 ----------------------
4939 1. Initialized the working variables associated with each extraction so that
4940 their saving and restoring doesn't refer to uninitialized store.
4942 2. Put dummy code into study.c in order to trick the optimizer of the IBM C
4943 compiler for OS/2 into generating correct code. Apparently IBM isn't going to
4944 fix the problem.
4946 3. Pcretest: the timing code wasn't using LOOPREPEAT for timing execution
4947 calls, and wasn't printing the correct value for compiling calls. Increased the
4948 default value of LOOPREPEAT, and the number of significant figures in the
4949 times.
4951 4. Changed "/bin/rm" in the Makefile to "-rm" so it works on Windows NT.
4953 5. Renamed "deftables" as "dftables" to get it down to 8 characters, to avoid
4954 a building problem on Windows NT with a FAT file system.
4957 Version 2.01 21-Oct-98
4958 ----------------------
4960 1. Changed the API for pcre_compile() to allow for the provision of a pointer
4961 to character tables built by pcre_maketables() in the current locale. If NULL
4962 is passed, the default tables are used.
4965 Version 2.00 24-Sep-98
4966 ----------------------
4968 1. Since the (>?) facility is in Perl 5.005, don't require PCRE_EXTRA to enable
4969 it any more.
4971 2. Allow quantification of (?>) groups, and make it work correctly.
4973 3. The first character computation wasn't working for (?>) groups.
4975 4. Correct the implementation of \Z (it is permitted to match on the \n at the
4976 end of the subject) and add 5.005's \z, which really does match only at the
4977 very end of the subject.
4979 5. Remove the \X "cut" facility; Perl doesn't have it, and (?> is neater.
4981 6. Remove the ability to specify CASELESS, MULTILINE, DOTALL, and
4982 DOLLAR_END_ONLY at runtime, to make it possible to implement the Perl 5.005
4983 localized options. All options to pcre_study() were also removed.
4985 7. Add other new features from 5.005:
4987 $(?<= positive lookbehind
4988 $(?<! negative lookbehind
4989 (?imsx-imsx) added the unsetting capability
4990 such a setting is global if at outer level; local otherwise
4991 (?imsx-imsx:) non-capturing groups with option setting
4992 (?(cond)re|re) conditional pattern matching
4994 A backreference to itself in a repeated group matches the previous
4995 captured string.
4997 8. General tidying up of studying (both automatic and via "study")
4998 consequential on the addition of new assertions.
5000 9. As in 5.005, unlimited repeated groups that could match an empty substring
5001 are no longer faulted at compile time. Instead, the loop is forcibly broken at
5002 runtime if any iteration does actually match an empty substring.
5004 10. Include the RunTest script in the distribution.
5006 11. Added tests from the Perl 5.005_02 distribution. This showed up a few
5007 discrepancies, some of which were old and were also with respect to 5.004. They
5008 have now been fixed.
5011 Version 1.09 28-Apr-98
5012 ----------------------
5014 1. A negated single character class followed by a quantifier with a minimum
5015 value of one (e.g. [^x]{1,6} ) was not compiled correctly. This could lead to
5016 program crashes, or just wrong answers. This did not apply to negated classes
5017 containing more than one character, or to minima other than one.
5020 Version 1.08 27-Mar-98
5021 ----------------------
5023 1. Add PCRE_UNGREEDY to invert the greediness of quantifiers.
5025 2. Add (?U) and (?X) to set PCRE_UNGREEDY and PCRE_EXTRA respectively. The
5026 latter must appear before anything that relies on it in the pattern.
5029 Version 1.07 16-Feb-98
5030 ----------------------
5032 1. A pattern such as /((a)*)*/ was not being diagnosed as in error (unlimited
5033 repeat of a potentially empty string).
5036 Version 1.06 23-Jan-98
5037 ----------------------
5039 1. Added Markus Oberhumer's little patches for C++.
5041 2. Literal strings longer than 255 characters were broken.
5044 Version 1.05 23-Dec-97
5045 ----------------------
5047 1. Negated character classes containing more than one character were failing if
5048 PCRE_CASELESS was set at run time.
5051 Version 1.04 19-Dec-97
5052 ----------------------
5054 1. Corrected the man page, where some "const" qualifiers had been omitted.
5056 2. Made debugging output print "{0,xxx}" instead of just "{,xxx}" to agree with
5057 input syntax.
5059 3. Fixed memory leak which occurred when a regex with back references was
5060 matched with an offsets vector that wasn't big enough. The temporary memory
5061 that is used in this case wasn't being freed if the match failed.
5063 4. Tidied pcretest to ensure it frees memory that it gets.
5065 5. Temporary memory was being obtained in the case where the passed offsets
5066 vector was exactly big enough.
5068 6. Corrected definition of offsetof() from change 5 below.
5070 7. I had screwed up change 6 below and broken the rules for the use of
5071 setjmp(). Now fixed.
5074 Version 1.03 18-Dec-97
5075 ----------------------
5077 1. A erroneous regex with a missing opening parenthesis was correctly
5078 diagnosed, but PCRE attempted to access brastack[-1], which could cause crashes
5079 on some systems.
5081 2. Replaced offsetof(real_pcre, code) by offsetof(real_pcre, code[0]) because
5082 it was reported that one broken compiler failed on the former because "code" is
5083 also an independent variable.
5085 3. The erroneous regex a[]b caused an array overrun reference.
5087 4. A regex ending with a one-character negative class (e.g. /[^k]$/) did not
5088 fail on data ending with that character. (It was going on too far, and checking
5089 the next character, typically a binary zero.) This was specific to the
5090 optimized code for single-character negative classes.
5092 5. Added a contributed patch from the TIN world which does the following:
5094 + Add an undef for memmove, in case the the system defines a macro for it.
5096 + Add a definition of offsetof(), in case there isn't one. (I don't know
5097 the reason behind this - offsetof() is part of the ANSI standard - but
5098 it does no harm).
5100 + Reduce the ifdef's in pcre.c using macro DPRINTF, thereby eliminating
5101 most of the places where whitespace preceded '#'. I have given up and
5102 allowed the remaining 2 cases to be at the margin.
5104 + Rename some variables in pcre to eliminate shadowing. This seems very
5105 pedantic, but does no harm, of course.
5107 6. Moved the call to setjmp() into its own function, to get rid of warnings
5108 from gcc -Wall, and avoided calling it at all unless PCRE_EXTRA is used.
5110 7. Constructs such as \d{8,} were compiling into the equivalent of
5111 \d{8}\d{0,65527} instead of \d{8}\d* which didn't make much difference to the
5112 outcome, but in this particular case used more store than had been allocated,
5113 which caused the bug to be discovered because it threw up an internal error.
5115 8. The debugging code in both pcre and pcretest for outputting the compiled
5116 form of a regex was going wrong in the case of back references followed by
5117 curly-bracketed repeats.
5120 Version 1.02 12-Dec-97
5121 ----------------------
5123 1. Typos in pcre.3 and comments in the source fixed.
5125 2. Applied a contributed patch to get rid of places where it used to remove
5126 'const' from variables, and fixed some signed/unsigned and uninitialized
5127 variable warnings.
5129 3. Added the "runtest" target to Makefile.
5131 4. Set default compiler flag to -O2 rather than just -O.
5134 Version 1.01 19-Nov-97
5135 ----------------------
5137 1. PCRE was failing to diagnose unlimited repeat of empty string for patterns
5138 like /([ab]*)*/, that is, for classes with more than one character in them.
5140 2. Likewise, it wasn't diagnosing patterns with "once-only" subpatterns, such
5141 as /((?>a*))*/ (a PCRE_EXTRA facility).
5144 Version 1.00 18-Nov-97
5145 ----------------------
5147 1. Added compile-time macros to support systems such as SunOS4 which don't have
5148 memmove() or strerror() but have other things that can be used instead.
5150 2. Arranged that "make clean" removes the executables.
5153 Version 0.99 27-Oct-97
5154 ----------------------
5156 1. Fixed bug in code for optimizing classes with only one character. It was
5157 initializing a 32-byte map regardless, which could cause it to run off the end
5158 of the memory it had got.
5160 2. Added, conditional on PCRE_EXTRA, the proposed (?>REGEX) construction.
5163 Version 0.98 22-Oct-97
5164 ----------------------
5166 1. Fixed bug in code for handling temporary memory usage when there are more
5167 back references than supplied space in the ovector. This could cause segfaults.
5170 Version 0.97 21-Oct-97
5171 ----------------------
5173 1. Added the \X "cut" facility, conditional on PCRE_EXTRA.
5175 2. Optimized negated single characters not to use a bit map.
5177 3. Brought error texts together as macro definitions; clarified some of them;
5178 fixed one that was wrong - it said "range out of order" when it meant "invalid
5179 escape sequence".
5181 4. Changed some char * arguments to const char *.
5183 5. Added PCRE_NOTBOL and PCRE_NOTEOL (from POSIX).
5185 6. Added the POSIX-style API wrapper in pcreposix.a and testing facilities in
5186 pcretest.
5189 Version 0.96 16-Oct-97
5190 ----------------------
5192 1. Added a simple "pgrep" utility to the distribution.
5194 2. Fixed an incompatibility with Perl: "{" is now treated as a normal character
5195 unless it appears in one of the precise forms "{ddd}", "{ddd,}", or "{ddd,ddd}"
5196 where "ddd" means "one or more decimal digits".
5198 3. Fixed serious bug. If a pattern had a back reference, but the call to
5199 pcre_exec() didn't supply a large enough ovector to record the related
5200 identifying subpattern, the match always failed. PCRE now remembers the number
5201 of the largest back reference, and gets some temporary memory in which to save
5202 the offsets during matching if necessary, in order to ensure that
5203 backreferences always work.
5205 4. Increased the compatibility with Perl in a number of ways:
5207 (a) . no longer matches \n by default; an option PCRE_DOTALL is provided
5208 to request this handling. The option can be set at compile or exec time.
5210 (b) $ matches before a terminating newline by default; an option
5211 PCRE_DOLLAR_ENDONLY is provided to override this (but not in multiline
5212 mode). The option can be set at compile or exec time.
5214 (c) The handling of \ followed by a digit other than 0 is now supposed to be
5215 the same as Perl's. If the decimal number it represents is less than 10
5216 or there aren't that many previous left capturing parentheses, an octal
5217 escape is read. Inside a character class, it's always an octal escape,
5218 even if it is a single digit.
5220 (d) An escaped but undefined alphabetic character is taken as a literal,
5221 unless PCRE_EXTRA is set. Currently this just reserves the remaining
5222 escapes.
5224 (e) {0} is now permitted. (The previous item is removed from the compiled
5225 pattern).
5227 5. Changed all the names of code files so that the basic parts are no longer
5228 than 10 characters, and abolished the teeny "globals.c" file.
5230 6. Changed the handling of character classes; they are now done with a 32-byte
5231 bit map always.
5233 7. Added the -d and /D options to pcretest to make it possible to look at the
5234 internals of compilation without having to recompile pcre.
5237 Version 0.95 23-Sep-97
5238 ----------------------
5240 1. Fixed bug in pre-pass concerning escaped "normal" characters such as \x5c or
5241 \x20 at the start of a run of normal characters. These were being treated as
5242 real characters, instead of the source characters being re-checked.
5245 Version 0.94 18-Sep-97
5246 ----------------------
5248 1. The functions are now thread-safe, with the caveat that the global variables
5249 containing pointers to malloc() and free() or alternative functions are the
5250 same for all threads.
5252 2. Get pcre_study() to generate a bitmap of initial characters for non-
5253 anchored patterns when this is possible, and use it if passed to pcre_exec().
5256 Version 0.93 15-Sep-97
5257 ----------------------
5259 1. /(b)|(:+)/ was computing an incorrect first character.
5261 2. Add pcre_study() to the API and the passing of pcre_extra to pcre_exec(),
5262 but not actually doing anything yet.
5264 3. Treat "-" characters in classes that cannot be part of ranges as literals,
5265 as Perl does (e.g. [-az] or [az-]).
5267 4. Set the anchored flag if a branch starts with .* or .*? because that tests
5268 all possible positions.
5270 5. Split up into different modules to avoid including unneeded functions in a
5271 compiled binary. However, compile and exec are still in one module. The "study"
5272 function is split off.
5274 6. The character tables are now in a separate module whose source is generated
5275 by an auxiliary program - but can then be edited by hand if required. There are
5276 now no calls to isalnum(), isspace(), isdigit(), isxdigit(), tolower() or
5277 toupper() in the code.
5279 7. Turn the malloc/free funtions variables into pcre_malloc and pcre_free and
5280 make them global. Abolish the function for setting them, as the caller can now
5281 set them directly.
5284 Version 0.92 11-Sep-97
5285 ----------------------
5287 1. A repeat with a fixed maximum and a minimum of 1 for an ordinary character
5288 (e.g. /a{1,3}/) was broken (I mis-optimized it).
5290 2. Caseless matching was not working in character classes if the characters in
5291 the pattern were in upper case.
5293 3. Make ranges like [W-c] work in the same way as Perl for caseless matching.
5295 4. Make PCRE_ANCHORED public and accept as a compile option.
5297 5. Add an options word to pcre_exec() and accept PCRE_ANCHORED and
5298 PCRE_CASELESS at run time. Add escapes \A and \I to pcretest to cause it to
5299 pass them.
5301 6. Give an error if bad option bits passed at compile or run time.
5303 7. Add PCRE_MULTILINE at compile and exec time, and (?m) as well. Add \M to
5304 pcretest to cause it to pass that flag.
5306 8. Add pcre_info(), to get the number of identifying subpatterns, the stored
5307 options, and the first character, if set.
5309 9. Recognize C+ or C{n,m} where n >= 1 as providing a fixed starting character.
5312 Version 0.91 10-Sep-97
5313 ----------------------
5315 1. PCRE was failing to diagnose unlimited repeats of subpatterns that could
5316 match the empty string as in /(a*)*/. It was looping and ultimately crashing.
5318 2. PCRE was looping on encountering an indefinitely repeated back reference to
5319 a subpattern that had matched an empty string, e.g. /(a|)\1*/. It now does what
5320 Perl does - treats the match as successful.
5322 ****


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

  ViewVC Help
Powered by ViewVC 1.1.5