/[pcre]/code/trunk/ChangeLog
ViewVC logotype

Diff of /code/trunk/ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 207 by ph10, Mon Aug 6 09:32:14 2007 UTC revision 1400 by ph10, Tue Nov 12 17:05:55 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.3 05-Jul-07  Version 8.34 xx-xxxx-201x
5    --------------------------
6    
7    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
8        executable memory. Patch inspired by Carsten Klein.
9    
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.
17    
18    3.  Make a small performance improvement in strlen16() and strlen32() in
19        pcretest.
20    
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.
24    
25    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
26    
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.
34    
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.
37    
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.
40    
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:
44    
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.
49    
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.
54    
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.
60    
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.
63    
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.
66    
67    13. Added the -T and -TM options to pcretest.
68    
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.
77    
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.
82    
83    16. Unicode character properties were updated from Unicode 6.3.0.
84    
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_POSSESS is added for testing purposes, and the -O and /O
89        options in pcretest are provided to set it. It can also be set by
90        (*NO_AUTO_POSSESS) at the start of a pattern.
91    
92    18. The character VT has been added to the set of characters that match \s and
93        are generally treated as white space, following this same change in Perl
94        5.18. There is now no difference between "Perl space" and "POSIX space".
95    
96    19. The code for checking named groups as conditions, either for being set or
97        for being recursed, has been refactored (this is related to 14 and 15
98        above). Processing unduplicated named groups should now be as fast at
99        numerical groups, and processing duplicated groups should be faster than
100        before.
101    
102    20. Two patches to the CMake build system, by Alexander Barkov:
103    
104          (1) Replace the "source" command by "." in CMakeLists.txt because
105              "source" is a bash-ism.
106    
107          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
108              without these the CMake build does not work on Solaris.
109    
110    21. Perl has changed its handling of \8 and \9. If there is no previously
111        encountered capturing group of those numbers, they are treated as the
112        literal characters 8 and 9 instead of a binary zero followed by the
113        literals. PCRE now does the same.
114    
115    22. Following Perl, added \o{} to specify codepoints in octal, making it
116        possible to specify values greater than 0777 and also making them
117        unambiguous.
118    
119    23. Perl now gives an error for missing closing braces after \x{... instead of
120        treating the string as literal. PCRE now does the same.
121    
122    24. RunTest used to grumble if an inappropriate test was selected explicitly,
123        but just skip it when running all tests. This make it awkward to run ranges
124        of tests when one of them was inappropriate. Now it just skips any
125        inappropriate tests, as it always did when running all tests.
126    
127    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
128        character types such as \d or \w, too many callouts were inserted, and the
129        data that they returned was rubbish.
130    
131    26. In UCP mode, \s was not matching two of the characters that Perl matches,
132        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
133        were matched by \h. The code has now been refactored so that the lists of
134        the horizontal and vertical whitespace characters used for \h and \v (which
135        are defined only in one place) are now also used for \s.
136    
137    27. Add JIT support for the 64 bit TileGX architecture.
138        Patch by Jiong Wang (Tilera Corporation).
139    
140    28. Possessive quantifiers for classes (both explicit and automatically
141        generated) now use special opcodes instead of wrapping in ONCE brackets.
142    
143    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
144        (because it's meaningless), this was not happening when PCRE_CASELESS was
145        set. Not wrong, but inefficient.
146    
147    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
148        properties for \w, \d, etc) is present in a test regex. Otherwise if the
149        test contains no characters greater than 255, Perl doesn't realise it
150        should be using Unicode semantics.
151    
152    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
153        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
154        does in Unicode mode.
155    
156    32. Added the "forbid" facility to pcretest so that putting tests into the
157        wrong test files can sometimes be quickly detected.
158    
159    33. There is now a limit (default 250) on the depth of nesting of parentheses.
160        This limit is imposed to control the amount of system stack used at compile
161        time. It can be changed at build time by --with-parens-nest-limit=xxx or
162        the equivalent in CMake.
163    
164    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
165        errors. Perl warns for these when in warning mode, but PCRE has no facility
166        for giving warnings.
167    
168    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
169        However, this was not working for (?!) because it is optimized to (*FAIL),
170        for which PCRE does not allow quantifiers. The optimization is now disabled
171        when a quantifier follows (?!). I can't see any use for this, but it makes
172        things uniform.
173    
174    36. Perl no longer allows group names to start with digits, so I have made this
175        change also in PCRE. It simplifies the code a bit.
176    
177    37. In extended mode, Perl ignores spaces before a + that indicates a
178        possessive quantifier. PCRE allowed a space before the quantifier, but not
179        before the possessive +. It now does.
180    
181    38. The use of \K (reset reported match start) within a repeated possessive
182        group such as (a\Kb)*+ was not working.
183    
184    40. Document that the same character tables must be used at compile time and
185        run time, and that the facility to pass tables to pcre_exec() and
186        pcre_dfa_exec() is for use only with saved/restored patterns.
187    
188    
189    Version 8.33 28-May-2013
190    ------------------------
191    
192    1.  Added 'U' to some constants that are compared to unsigned integers, to
193        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
194        variables that are added to signed variables, to ensure the result is
195        signed and can be negated.
196    
197    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
198        CMake config files.
199    
200    3.  Revise the creation of config.h.generic so that all boolean macros are
201        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
202        overriding via -D on the command line possible.
203    
204    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
205        to unsigned int is reported to make a quite noticeable speed difference in
206        a specific Windows environment. Testing on Linux did also appear to show
207        some benefit (and it is clearly not harmful). Also fixed the definition of
208        Xop which should be unsigned.
209    
210    5.  Related to (4), changing the definition of the intermediate variable cc
211        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
212        improvements.
213    
214    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
215        unnecessary spaces.
216    
217    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
218        and later.
219    
220    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
221    
222    9.  Optimizing fast_forward_start_bits in JIT.
223    
224    10. Adding support for callouts in JIT, and fixing some issues revealed
225        during this work. Namely:
226    
227        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
228    
229        (b) Minimum length was not checked before the matching is started.
230    
231    11. The value of capture_last that is passed to callouts was incorrect in some
232        cases when there was a capture on one path that was subsequently abandoned
233        after a backtrack. Also, the capture_last value is now reset after a
234        recursion, since all captures are also reset in this case.
235    
236    12. The interpreter no longer returns the "too many substrings" error in the
237        case when an overflowing capture is in a branch that is subsequently
238        abandoned after a backtrack.
239    
240    13. In the pathological case when an offset vector of size 2 is used, pcretest
241        now prints out the matched string after a yield of 0 or 1.
242    
243    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
244        Only supported by the JIT compiler at the moment.
245    
246    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
247    
248    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
249        offset of the starting point of the matching process, provided the offsets
250        vector is large enough.
251    
252    17. The \A escape now records a lookbehind value of 1, though its execution
253        does not actually inspect the previous character. This is to ensure that,
254        in partial multi-segment matching, at least one character from the old
255        segment is retained when a new segment is processed. Otherwise, if there
256        are no lookbehinds in the pattern, \A might match incorrectly at the start
257        of a new segment.
258    
259    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
260    
261    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
262        gives some modest performance improvement in 8-bit mode.
263    
264    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
265        be expressed in certain programming languages using Universal Character
266        Names.
267    
268    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
269        which points out that "non characters" are not "characters that may not
270        appear in Unicode strings" but rather "characters that are reserved for
271        internal use and have only local meaning".
272    
273    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
274        there was a conditional group that depended on an assertion, if the
275        assertion was false, the callout that immediately followed the alternation
276        in the condition was skipped when pcre_exec() was used for matching.
277    
278    23. Allow an explicit callout to be inserted before an assertion that is the
279        condition for a conditional group, for compatibility with automatic
280        callouts, which always insert a callout at this point.
281    
282    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
283        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
284    
285    25. (*PRUNE) is now supported by the JIT compiler.
286    
287    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
288    
289    27. Fix the case where there are two or more SKIPs with arguments that may be
290        ignored.
291    
292    28. (*SKIP) is now supported by the JIT compiler.
293    
294    29. (*THEN) is now supported by the JIT compiler.
295    
296    30. Update RunTest with additional test selector options.
297    
298    31. The way PCRE handles backtracking verbs has been changed in two ways.
299    
300        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
301        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
302        backtracking. In some cases this makes it more Perl-compatible, but Perl's
303        rather obscure rules do not always do the same thing.
304    
305        (2) Previously, backtracking verbs were confined within assertions. This is
306        no longer the case for positive assertions, except for (*ACCEPT). Again,
307        this sometimes improves Perl compatibility, and sometimes does not.
308    
309    32. A number of tests that were in test 2 because Perl did things differently
310        have been moved to test 1, because either Perl or PCRE has changed, and
311        these tests are now compatible.
312    
313    32. Backtracking control verbs are now handled in the same way in JIT and
314        interpreter.
315    
316    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
317        contained a forward subroutine reference caused a compile error.
318    
319    34. Auto-detect and optimize limited repetitions in JIT.
320    
321    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
322        blocking (*UTF) etc.
323    
324    36. In the interpreter, maximizing pattern repetitions for characters and
325        character types now use tail recursion, which reduces stack usage.
326    
327    37. The value of the max lookbehind was not correctly preserved if a compiled
328        and saved regex was reloaded on a host of different endianness.
329    
330    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
331        of the compiled pattern block, expand the flags field from 16 to 32 bits
332        because it was almost full.
333    
334    39. Try madvise first before posix_madvise.
335    
336    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
337        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
338        original change locked out all matches of empty strings. This has been
339        changed so that one match of an empty string per line is recognized.
340        Subsequent searches on the same line (for colouring or for --only-matching,
341        for example) do not recognize empty strings.
342    
343    41. Applied a user patch to fix a number of spelling mistakes in comments.
344    
345    42. Data lines longer than 65536 caused pcretest to crash.
346    
347    43. Clarified the data type for length and startoffset arguments for pcre_exec
348        and pcre_dfa_exec in the function-specific man pages, where they were
349        explicitly stated to be in bytes, never having been updated. I also added
350        some clarification to the pcreapi man page.
351    
352    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
353        a segmentation fault.
354    
355    
356    Version 8.32 30-November-2012
357    -----------------------------
358    
359    1.  Improved JIT compiler optimizations for first character search and single
360        character iterators.
361    
362    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
363        Patch by Daniel Richard G.
364    
365    3.  Single character iterator optimizations in the JIT compiler.
366    
367    4.  Improved JIT compiler optimizations for character ranges.
368    
369    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
370        Reported by Giuseppe D'Angelo.
371    
372    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
373        of a line, was being set incorrectly in cases where .* appeared inside
374        atomic brackets at the start of a pattern, or where there was a subsequent
375        *PRUNE or *SKIP.
376    
377    7.  Improved instruction cache flush for POWER/PowerPC.
378        Patch by Daniel Richard G.
379    
380    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
381        grep:
382    
383        (a) There is now no limit to the number of patterns to be matched.
384    
385        (b) An error is given if a pattern is too long.
386    
387        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
388            are now supported.
389    
390        (d) --exclude-from and --include-from (multiple use) have been added.
391    
392        (e) Exclusions and inclusions now apply to all files and directories, not
393            just to those obtained from scanning a directory recursively.
394    
395        (f) Multiple uses of -f and --file-list are now supported.
396    
397        (g) In a Windows environment, the default for -d has been changed from
398            "read" (the GNU grep default) to "skip", because otherwise the presence
399            of a directory in the file list provokes an error.
400    
401        (h) The documentation has been revised and clarified in places.
402    
403    9.  Improve the matching speed of capturing brackets.
404    
405    10. Changed the meaning of \X so that it now matches a Unicode extended
406        grapheme cluster.
407    
408    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
409        out POSIX threads when JIT support is configured.
410    
411    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
412    
413    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
414        structure could go wrong in environments where size_t is not the same size
415        as int.
416    
417    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
418        set.
419    
420    15. The EBCDIC support had decayed; later updates to the code had included
421        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
422        general tidy up of EBCDIC-related issues, and the documentation was also
423        not quite right. There is now a test that can be run on ASCII systems to
424        check some of the EBCDIC-related things (but is it not a full test).
425    
426    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
427        in a small tidy to the code.
428    
429    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
430    
431    18. If the --only-matching (-o) option in pcregrep is specified multiple
432        times, each one causes appropriate output. For example, -o1 -o2 outputs the
433        substrings matched by the 1st and 2nd capturing parentheses. A separating
434        string can be specified by --om-separator (default empty).
435    
436    19. Improving the first n character searches.
437    
438    20. Turn case lists for horizontal and vertical white space into macros so that
439        they are defined only once.
440    
441    21. This set of changes together give more compatible Unicode case-folding
442        behaviour for characters that have more than one other case when UCP
443        support is available.
444    
445        (a) The Unicode property table now has offsets into a new table of sets of
446            three or more characters that are case-equivalent. The MultiStage2.py
447            script that generates these tables (the pcre_ucd.c file) now scans
448            CaseFolding.txt instead of UnicodeData.txt for character case
449            information.
450    
451        (b) The code for adding characters or ranges of characters to a character
452            class has been abstracted into a generalized function that also handles
453            case-independence. In UTF-mode with UCP support, this uses the new data
454            to handle characters with more than one other case.
455    
456        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
457            whose other case is greater than 256 are now correctly matched
458            caselessly. Previously, the high codepoint matched the low one, but not
459            vice versa.
460    
461        (d) The processing of \h, \H, \v, and \ in character classes now makes use
462            of the new class addition function, using character lists defined as
463            macros alongside the case definitions of 20 above.
464    
465        (e) Caseless back references now work with characters that have more than
466            one other case.
467    
468        (f) General caseless matching of characters with more than one other case
469            is supported.
470    
471    22. Unicode character properties were updated from Unicode 6.2.0
472    
473    23. Improved CMake support under Windows. Patch by Daniel Richard G.
474    
475    24. Add support for 32-bit character strings, and UTF-32
476    
477    25. Major JIT compiler update (code refactoring and bugfixing).
478        Experimental Sparc 32 support is added.
479    
480    26. Applied a modified version of Daniel Richard G's patch to create
481        pcre.h.generic and config.h.generic by "make" instead of in the
482        PrepareRelease script.
483    
484    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
485        pcre_compile.c when checking for a zero character.
486    
487    28. Introducing a native interface for JIT. Through this interface, the compiled
488        machine code can be directly executed. The purpose of this interface is to
489        provide fast pattern matching, so several sanity checks are not performed.
490        However, feature tests are still performed. The new interface provides
491        1.4x speedup compared to the old one.
492    
493    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
494        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
495        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
496    
497    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
498        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
499        the "old" RFC 2279). Instead, it ended up passing a negative length to
500        pcre_exec().
501    
502    31. Add support for GCC's visibility feature to hide internal functions.
503    
504    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
505        "unknown -C option" after outputting 0 or 1.
506    
507    33. There is now support for generating a code coverage report for the test
508        suite in environments where gcc is the compiler and lcov is installed. This
509        is mainly for the benefit of the developers.
510    
511    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
512        unaddressable using valgrind annotations, allowing valgrind to detect
513        invalid memory accesses. This is mainly for the benefit of the developers.
514    
515    25. (*UTF) can now be used to start a pattern in any of the three libraries.
516    
517    26. Give configure error if --enable-cpp but no C++ compiler found.
518    
519    
520    Version 8.31 06-July-2012
521    -------------------------
522    
523    1.  Fixing a wrong JIT test case and some compiler warnings.
524    
525    2.  Removed a bashism from the RunTest script.
526    
527    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
528        to unsigned type, result still unsigned" that was given by an MS compiler
529        on encountering the code "-sizeof(xxx)".
530    
531    4.  Partial matching support is added to the JIT compiler.
532    
533    5.  Fixed several bugs concerned with partial matching of items that consist
534        of more than one character:
535    
536        (a) /^(..)\1/ did not partially match "aba" because checking references was
537            done on an "all or nothing" basis. This also applied to repeated
538            references.
539    
540        (b) \R did not give a hard partial match if \r was found at the end of the
541            subject.
542    
543        (c) \X did not give a hard partial match after matching one or more
544            characters at the end of the subject.
545    
546        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
547            a partial match for the string "\r".
548    
549        (e) When newline was set to CRLF, the metacharacter "." did not recognize
550            a partial match for a CR character at the end of the subject string.
551    
552    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
553        running pcretest, the text "(JIT)" added to the output whenever JIT is
554        actually used to run the match.
555    
556    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
557        or /S+[+] with a digit between 1 and 7.
558    
559    8.  OP_NOT now supports any UTF character not just single-byte ones.
560    
561    9.  (*MARK) control verb is now supported by the JIT compiler.
562    
563    10. The command "./RunTest list" lists the available tests without actually
564        running any of them. (Because I keep forgetting what they all are.)
565    
566    11. Add PCRE_INFO_MAXLOOKBEHIND.
567    
568    12. Applied a (slightly modified) user-supplied patch that improves performance
569        when the heap is used for recursion (compiled with --disable-stack-for-
570        recursion). Instead of malloc and free for each heap frame each time a
571        logical recursion happens, frames are retained on a chain and re-used where
572        possible. This sometimes gives as much as 30% improvement.
573    
574    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
575        call.
576    
577    14. As documented, (*COMMIT) is now confined to within a positive assertion.
578    
579    15. It is now possible to link pcretest with libedit as an alternative to
580        libreadline.
581    
582    16. (*COMMIT) control verb is now supported by the JIT compiler.
583    
584    17. The Unicode data tables have been updated to Unicode 6.1.0.
585    
586    18. Added --file-list option to pcregrep.
587    
588    19. Added binary file support to pcregrep, including the -a, --binary-files,
589        -I, and --text options.
590    
591    20. The madvise function is renamed for posix_madvise for QNX compatibility
592        reasons. Fixed by Giuseppe D'Angelo.
593    
594    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
595    
596    22. Changed the option for creating symbolic links for 16-bit man pages from
597        -s to -sf so that re-installing does not cause issues.
598    
599    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
600    
601    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
602        matches in certain environments (the workspace was not being correctly
603        retained). Also added to pcre_dfa_exec() a simple plausibility check on
604        some of the workspace data at the beginning of a restart.
605    
606    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
607        was not doing so when it should - probably a typo introduced by SVN 528
608        (change 8.10/14).
609    
610    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
611        \w+ when the character tables indicated that \x{c4} was a word character.
612        There were several related cases, all because the tests for doing a table
613        lookup were testing for characters less than 127 instead of 255.
614    
615    27. If a pattern contains capturing parentheses that are not used in a match,
616        their slots in the ovector are set to -1. For those that are higher than
617        any matched groups, this happens at the end of processing. In the case when
618        there were back references that the ovector was too small to contain
619        (causing temporary malloc'd memory to be used during matching), and the
620        highest capturing number was not used, memory off the end of the ovector
621        was incorrectly being set to -1. (It was using the size of the temporary
622        memory instead of the true size.)
623    
624    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
625        ovector size, it uses memory at the end of the block that it has got.
626    
627    29. Check for an overlong MARK name and give an error at compile time. The
628        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
629    
630    30. JIT compiler update.
631    
632    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
633        Rill for the patch.
634    
635    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
636    
637    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
638    
639    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
640        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
641        was enabled.
642    
643    35. Improve JIT code generation for greedy plus quantifier.
644    
645    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
646        1 to "aa" instead of to an empty string. The bug affected repeated groups
647        that could potentially match an empty string.
648    
649    37. Optimizing single character iterators in JIT.
650    
651    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
652        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
653        codepoints that are too big for the mode are faulted, and in a UTF mode,
654        disallowed codepoints are also faulted.
655    
656    39. If PCRE was compiled with UTF support, in three places in the DFA
657        matcher there was code that should only have been obeyed in UTF mode, but
658        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
659        processing when bytes with values greater than 127 were present. In 16-bit
660        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
661        both cases the values are those that cannot be the first data item in a UTF
662        character. The three items that might have provoked this were recursions,
663        possessively repeated groups, and atomic groups.
664    
665    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
666        and pcregrep, because some OS require shared objects to be explicitly
667        passed to ld, causing the link step to fail if they are not.
668    
669    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
670        mode, patterns that started with \h* or \R* might be incorrectly matched.
671    
672    
673    Version 8.30 04-February-2012
674    -----------------------------
675    
676    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
677        name is defined in ctype.h.
678    
679    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
680        only in quite long subpatterns.
681    
682    3.  Removed the function pcre_info(), which has been obsolete and deprecated
683        since it was replaced by pcre_fullinfo() in February 2000.
684    
685    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
686        match a (*MARK), and the match failed at the start of the subject, a
687        reference to memory before the start of the subject could occur. This bug
688        was introduced by fix 17 of release 8.21.
689    
690    5.  A reference to an unset group with zero minimum repetition was giving
691        totally wrong answers (in non-JavaScript-compatibility mode). For example,
692        /(another)?(\1?)test/ matched against "hello world test". This bug was
693        introduced in release 8.13.
694    
695    6.  Add support for 16-bit character strings (a large amount of work involving
696        many changes and refactorings).
697    
698    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
699        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
700        from a file.
701    
702    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
703        rounding is not applied in this particular case).
704    
705    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
706        if they appear, or are escaped, in patterns.
707    
708    10. Get rid of a number of -Wunused-but-set-variable warnings.
709    
710    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
711        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
712        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
713        also returns the mark "x". This bug applied to capturing parentheses,
714        non-capturing parentheses, and atomic parentheses. It also applied to some
715        assertions.
716    
717    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
718        information out of configure.ac instead of relying on pcre.h.generic, which
719        is not stored in the repository.
720    
721    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
722        -lreadline.
723    
724    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
725    
726    15. Applied Graycode's patch to put the top-level frame on the stack rather
727        than the heap when not using the stack for recursion. This gives a
728        performance improvement in many cases when recursion is not deep.
729    
730    16. Experimental code added to "pcretest -C" to output the stack frame size.
731    
732    
733    Version 8.21 12-Dec-2011
734    ------------------------
735    
736    1.  Updating the JIT compiler.
737    
738    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
739        are added as well.
740    
741    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
742        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
743        calling _pcre_jit_exec. Some extra comments are added.
744    
745    4.  (*MARK) settings inside atomic groups that do not contain any capturing
746        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
747        was introduced by change 18 for 8.20.
748    
749    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
750        ECMA-262 standard.
751    
752    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
753        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
754        This bug was probably introduced by change 9 of 8.13.
755    
756    7.  While fixing 6 above, I noticed that a number of other items were being
757        incorrectly rejected as "not fixed length". This arose partly because newer
758        opcodes had not been added to the fixed-length checking code. I have (a)
759        corrected the bug and added tests for these items, and (b) arranged for an
760        error to occur if an unknown opcode is encountered while checking for fixed
761        length instead of just assuming "not fixed length". The items that were
762        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
763        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
764        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
765    
766    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
767        being incorrectly compiled and would have given unpredicatble results.
768    
769    9.  A possessively repeated subpattern with minimum repeat count greater than
770        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
771        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
772        the first (A) could occur when it should not.
773    
774    10. Add a cast and remove a redundant test from the code.
775    
776    11. JIT should use pcre_malloc/pcre_free for allocation.
777    
778    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
779        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
780        is anything other than /usr, -L is still shown).
781    
782    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
783    
784    14. Perl does not support \N without a following name in a [] class; PCRE now
785        also gives an error.
786    
787    15. If a forward reference was repeated with an upper limit of around 2000,
788        it caused the error "internal error: overran compiling workspace". The
789        maximum number of forward references (including repeats) was limited by the
790        internal workspace, and dependent on the LINK_SIZE. The code has been
791        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
792        the default depends on LINK_SIZE. There is a new upper limit (for safety)
793        of around 200,000 forward references. While doing this, I also speeded up
794        the filling in of repeated forward references.
795    
796    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
797        incorrectly expecting the subject to contain another "a" after the start.
798    
799    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
800        in the match, the SKIP should be ignored. This was not happening; instead
801        the SKIP was being treated as NOMATCH. For patterns such as
802        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
803        tested.
804    
805    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
806        now much more compatible with Perl, in particular in cases where the result
807        is a non-match for a non-anchored pattern. For example, if
808        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
809        "m", where previously it did not return a name. A side effect of this
810        change is that for partial matches, the last encountered mark name is
811        returned, as for non matches. A number of tests that were previously not
812        Perl-compatible have been moved into the Perl-compatible test files. The
813        refactoring has had the pleasing side effect of removing one argument from
814        the match() function, thus reducing its stack requirements.
815    
816    19. If the /S+ option was used in pcretest to study a pattern using JIT,
817        subsequent uses of /S (without +) incorrectly behaved like /S+.
818    
819    21. Retrieve executable code size support for the JIT compiler and fixing
820        some warnings.
821    
822    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
823        not work when the shorter character appeared right at the end of the
824        subject string.
825    
826    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
827        systems.
828    
829    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
830        output it when the /M option is used in pcretest.
831    
832    25. The CheckMan script was not being included in the distribution. Also, added
833        an explicit "perl" to run Perl scripts from the PrepareRelease script
834        because this is reportedly needed in Windows.
835    
836    26. If study data was being save in a file and studying had not found a set of
837        "starts with" bytes for the pattern, the data written to the file (though
838        never used) was taken from uninitialized memory and so caused valgrind to
839        complain.
840    
841    27. Updated RunTest.bat as provided by Sheri Pierce.
842    
843    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
844    
845    29. Computation of memory usage for the table of capturing group names was
846        giving an unnecessarily large value.
847    
848    
849    Version 8.20 21-Oct-2011
850    ------------------------
851    
852    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
853        a POSIX class. After further experiments with Perl, which convinced me that
854        Perl has bugs and confusions, a closing square bracket is no longer allowed
855        in a POSIX name. This bug also affected patterns with classes that started
856        with full stops.
857    
858    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
859        captured substring, but while checking the failing first alternative,
860        substring 1 is temporarily captured. If the output vector supplied to
861        pcre_exec() was not big enough for this capture, the yield of the function
862        was still zero ("insufficient space for captured substrings"). This cannot
863        be totally fixed without adding another stack variable, which seems a lot
864        of expense for a edge case. However, I have improved the situation in cases
865        such as /(a)(b)x|abc/ matched against "abc", where the return code
866        indicates that fewer than the maximum number of slots in the ovector have
867        been set.
868    
869    3.  Related to (2) above: when there are more back references in a pattern than
870        slots in the output vector, pcre_exec() uses temporary memory during
871        matching, and copies in the captures as far as possible afterwards. It was
872        using the entire output vector, but this conflicts with the specification
873        that only 2/3 is used for passing back captured substrings. Now it uses
874        only the first 2/3, for compatibility. This is, of course, another edge
875        case.
876    
877    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
878        main code base, and can be used by building with --enable-jit. When this is
879        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
880        runtime --no-jit option is given.
881    
882    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
883        ovector, the return from the function was zero, implying that there were
884        other matches that did not fit. The correct "exactly full" value is now
885        returned.
886    
887    6.  If a subpattern that was called recursively or as a subroutine contained
888        (*PRUNE) or any other control that caused it to give a non-standard return,
889        invalid errors such as "Error -26 (nested recursion at the same subject
890        position)" or even infinite loops could occur.
891    
892    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
893        computing the minimum length on reaching *ACCEPT, and so ended up with the
894        wrong value of 1 rather than 0. Further investigation indicates that
895        computing a minimum subject length in the presence of *ACCEPT is difficult
896        (think back references, subroutine calls), and so I have changed the code
897        so that no minimum is registered for a pattern that contains *ACCEPT.
898    
899    8.  If (*THEN) was present in the first (true) branch of a conditional group,
900        it was not handled as intended. [But see 16 below.]
901    
902    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
903        Sheri Pierce.
904    
905    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
906        the first byte in a match must be "a".
907    
908    11. Change 17 for 8.13 increased the recursion depth for patterns like
909        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
910        pattern contains any instances of (*THEN). If it does not, the old
911        optimizations are restored. It would be nice to do this on a per-group
912        basis, but at the moment that is not feasible.
913    
914    12. In some environments, the output of pcretest -C is CRLF terminated. This
915        broke RunTest's code that checks for the link size. A single white space
916        character after the value is now allowed for.
917    
918    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
919        For "fr", it uses the Windows-specific input and output files.
920    
921    14. If (*THEN) appeared in a group that was called recursively or as a
922        subroutine, it did not work as intended. [But see next item.]
923    
924    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
925        pattern fragments (but not containing any | characters). If A and B are
926        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
927        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
928        D. In other words, Perl considers parentheses that do not contain any |
929        characters to be part of a surrounding alternative, whereas PCRE was
930        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
931        differently. PCRE now behaves in the same way as Perl, except in the case
932        of subroutine/recursion calls such as (?1) which have in any case always
933        been different (but PCRE had them first :-).
934    
935    16. Related to 15 above: Perl does not treat the | in a conditional group as
936        creating alternatives. Such a group is treated in the same way as an
937        ordinary group without any | characters when processing (*THEN). PCRE has
938        been changed to match Perl's behaviour.
939    
940    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
941        RunGrepTest script failed.
942    
943    18. Change 22 for version 13 caused atomic groups to use more stack. This is
944        inevitable for groups that contain captures, but it can lead to a lot of
945        stack use in large patterns. The old behaviour has been restored for atomic
946        groups that do not contain any capturing parentheses.
947    
948    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
949        suppress the check for a minimum subject length at run time. (If it was
950        given to pcre_exec() or pcre_dfa_exec() it did work.)
951    
952    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
953        fail to work when decoding hex characters in data strings in EBCDIC
954        environments.
955    
956    21. It appears that in at least one Mac OS environment, the isxdigit() function
957        is implemented as a macro that evaluates to its argument more than once,
958        contravening the C 90 Standard (I haven't checked a later standard). There
959        was an instance in pcretest which caused it to go wrong when processing
960        \x{...} escapes in subject strings. The has been rewritten to avoid using
961        things like p++ in the argument of isxdigit().
962    
963    
964    Version 8.13 16-Aug-2011
965    ------------------------
966    
967    1.  The Unicode data tables have been updated to Unicode 6.0.0.
968    
969    2.  Two minor typos in pcre_internal.h have been fixed.
970    
971    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
972        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
973        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
974    
975    4.  There were a number of related bugs in the code for matching backrefences
976        caselessly in UTF-8 mode when codes for the characters concerned were
977        different numbers of bytes. For example, U+023A and U+2C65 are an upper
978        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
979        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
980        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
981        2-byte code at the end of the subject (it thought there wasn't enough data
982        left).
983    
984    5.  Comprehensive information about what went wrong is now returned by
985        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
986        as the output vector has at least 2 elements. The offset of the start of
987        the failing character and a reason code are placed in the vector.
988    
989    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
990        now returned is for the first byte of the failing character, instead of the
991        last byte inspected. This is an incompatible change, but I hope it is small
992        enough not to be a problem. It makes the returned offset consistent with
993        pcre_exec() and pcre_dfa_exec().
994    
995    7.  pcretest now gives a text phrase as well as the error number when
996        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
997        failure, the offset and reason code are output.
998    
999    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1000        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1001        back over a single character (\n). This seems wrong (because it treated the
1002        two characters as a single entity when going forwards), conflicts with the
1003        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1004        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1005        has been changed.
1006    
1007    9.  Some internal refactoring has changed the processing so that the handling
1008        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1009        time (the PCRE_DOTALL option was changed this way some time ago: version
1010        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1011        which was always a bit of a fudge. It also means that there is one less
1012        argument for the match() function, which reduces its stack requirements
1013        slightly. This change also fixes an incompatibility with Perl: the pattern
1014        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1015    
1016    10. More internal refactoring has drastically reduced the number of recursive
1017        calls to match() for possessively repeated groups such as (abc)++ when
1018        using pcre_exec().
1019    
1020    11. While implementing 10, a number of bugs in the handling of groups were
1021        discovered and fixed:
1022    
1023        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1024        (a|)*(?1) gave a compile-time internal error.
1025        ((a|)+)+  did not notice that the outer group could match an empty string.
1026        (^a|^)+   was not marked as anchored.
1027        (.*a|.*)+ was not marked as matching at start or after a newline.
1028    
1029    12. Yet more internal refactoring has removed another argument from the match()
1030        function. Special calls to this function are now indicated by setting a
1031        value in a variable in the "match data" data block.
1032    
1033    13. Be more explicit in pcre_study() instead of relying on "default" for
1034        opcodes that mean there is no starting character; this means that when new
1035        ones are added and accidentally left out of pcre_study(), testing should
1036        pick them up.
1037    
1038    14. The -s option of pcretest has been documented for ages as being an old
1039        synonym of -m (show memory usage). I have changed it to mean "force study
1040        for every regex", that is, assume /S for every regex. This is similar to -i
1041        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1042        using it. It makes it easier to run collections of tests with and without
1043        study enabled, and thereby test pcre_study() more easily. All the standard
1044        tests are now run with and without -s (but some patterns can be marked as
1045        "never study" - see 20 below).
1046    
1047    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1048        restoration of the capturing data to the outer values was not happening
1049        correctly.
1050    
1051    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1052        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1053        pattern had matched an empty string, and so incorrectly returned a no
1054        match.
1055    
1056    17. There was optimizing code for the last branch of non-capturing parentheses,
1057        and also for the obeyed branch of a conditional subexpression, which used
1058        tail recursion to cut down on stack usage. Unfortunately, now that there is
1059        the possibility of (*THEN) occurring in these branches, tail recursion is
1060        no longer possible because the return has to be checked for (*THEN). These
1061        two optimizations have therefore been removed. [But see 8.20/11 above.]
1062    
1063    18. If a pattern containing \R was studied, it was assumed that \R always
1064        matched two bytes, thus causing the minimum subject length to be
1065        incorrectly computed because \R can also match just one byte.
1066    
1067    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1068        was incorrectly computed.
1069    
1070    20. If /S is present twice on a test pattern in pcretest input, it now
1071        *disables* studying, thereby overriding the use of -s on the command line
1072        (see 14 above). This is necessary for one or two tests to keep the output
1073        identical in both cases.
1074    
1075    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1076        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1077    
1078    22. When an atomic group that contained a capturing parenthesis was
1079        successfully matched, but the branch in which it appeared failed, the
1080        capturing was not being forgotten if a higher numbered group was later
1081        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1082        group 1 to "a", when in fact it should be unset. This applied to multi-
1083        branched capturing and non-capturing groups, repeated or not, and also to
1084        positive assertions (capturing in negative assertions does not happen
1085        in PCRE) and also to nested atomic groups.
1086    
1087    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1088        subject after a captured substring, to make it easier to tell which of a
1089        number of identical substrings has been captured.
1090    
1091    24. The way atomic groups are processed by pcre_exec() has been changed so that
1092        if they are repeated, backtracking one repetition now resets captured
1093        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1094        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1095        "aaa". Previously, it would have been "a". As part of this code
1096        refactoring, the way recursive calls are handled has also been changed.
1097    
1098    25. If an assertion condition captured any substrings, they were not passed
1099        back unless some other capturing happened later. For example, if
1100        (?(?=(a))a) was matched against "a", no capturing was returned.
1101    
1102    26. When studying a pattern that contained subroutine calls or assertions,
1103        the code for finding the minimum length of a possible match was handling
1104        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1105        group 1 called group 2 while simultaneously a separate group 2 called group
1106        1). A stack overflow occurred in this case. I have fixed this by limiting
1107        the recursion depth to 10.
1108    
1109    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1110        Fortmann. This supports explicit test numbers on the command line, and has
1111        argument validation and error reporting.
1112    
1113    28. An instance of \X with an unlimited repeat could fail if at any point the
1114        first character it looked at was a mark character.
1115    
1116    29. Some minor code refactoring concerning Unicode properties and scripts
1117        should reduce the stack requirement of match() slightly.
1118    
1119    30. Added the '=' option to pcretest to check the setting of unused capturing
1120        slots at the end of the pattern, which are documented as being -1, but are
1121        not included in the return count.
1122    
1123    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1124        compiled something random. Now it gives a compile-time error (as does
1125        Perl).
1126    
1127    32. A *MARK encountered during the processing of a positive assertion is now
1128        recorded and passed back (compatible with Perl).
1129    
1130    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1131        had alternative anchored branches, the search for a second match in a line
1132        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1133        matched the line "0102" twice. The same bug affected patterns that started
1134        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1135        twice.
1136    
1137    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1138        does, and because of capturing effects, quantifying parenthesized
1139        assertions may at times be useful. Quantifiers are now allowed for
1140        parenthesized assertions.
1141    
1142    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1143    
1144    36. \g was being checked for fancy things in a character class, when it should
1145        just be a literal "g".
1146    
1147    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1148        appearance of a nested POSIX class supersedes an apparent external class.
1149        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1150        unescaped square brackets may also appear as part of class names. For
1151        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1152        more like Perl. (But see 8.20/1 above.)
1153    
1154    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1155        was because it thought it was \N{name}, which is not supported).
1156    
1157    39. Add minix to OS list not supporting the -S option in pcretest.
1158    
1159    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1160        cannot analyze patterns in sufficient detail to catch mutual recursions
1161        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1162        subgroup is called recursively as a subpattern for a second time at the
1163        same position in the subject string. In previous releases this might have
1164        been caught by the recursion limit, or it might have run out of stack.
1165    
1166    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1167        happen only once. PCRE was, however incorrectly giving a compile time error
1168        "recursive call could loop indefinitely" because it cannot analyze the
1169        pattern in sufficient detail. The compile time test no longer happens when
1170        PCRE is compiling a conditional subpattern, but actual runaway loops are
1171        now caught at runtime (see 40 above).
1172    
1173    42. It seems that Perl allows any characters other than a closing parenthesis
1174        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1175        has been changed to be the same.
1176    
1177    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1178        as not to get warnings when autogen.sh is called. Also changed
1179        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1180    
1181    44. To help people who use pcregrep to scan files containing exceedingly long
1182        lines, the following changes have been made:
1183    
1184        (a) The default value of the buffer size parameter has been increased from
1185            8K to 20K. (The actual buffer used is three times this size.)
1186    
1187        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1188            PCRE is built.
1189    
1190        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1191            to be set at run time.
1192    
1193        (d) Numerical values in pcregrep options can be followed by K or M, for
1194            example --buffer-size=50K.
1195    
1196        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1197            given and the return code is set to 2.
1198    
1199    45. Add a pointer to the latest mark to the callout data block.
1200    
1201    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1202        partial match of an empty string instead of no match. This was specific to
1203        the use of ".".
1204    
1205    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1206        complete match instead of a partial match. This bug was dependent on both
1207        the PCRE_UTF8 and PCRE_DOTALL options being set.
1208    
1209    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1210        starting byte set, because \b was not being ignored.
1211    
1212    
1213    Version 8.12 15-Jan-2011
1214    ------------------------
1215    
1216    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1217        checks for such things as part of the documentation building process.
1218    
1219    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1220        --match-limit and --recursion-limit options (added for 8.11). In
1221        particular, this made one of the standard tests fail. (The integer value
1222        went into the wrong half of a long int.)
1223    
1224    3.  If the --colour option was given to pcregrep with -v (invert match), it
1225        did strange things, either producing crazy output, or crashing. It should,
1226        of course, ignore a request for colour when reporting lines that do not
1227        match.
1228    
1229    4.  Another pcregrep bug caused similar problems if --colour was specified with
1230        -M (multiline) and the pattern match finished with a line ending.
1231    
1232    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1233        matched in multiline mode, the following line was shown as part of the
1234        match. This seems wrong, so I have changed it.
1235    
1236    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1237        the check for further matches in the same line (so they could be coloured)
1238        to overrun the end of the current line. If another match was found, it was
1239        incorrectly shown (and then shown again when found in the next line).
1240    
1241    7.  If pcregrep was compiled under Windows, there was a reference to the
1242        function pcregrep_exit() before it was defined. I am assuming this was
1243        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1244        reported by a user. I've moved the definition above the reference.
1245    
1246    
1247    Version 8.11 10-Dec-2010
1248    ------------------------
1249    
1250    1.  (*THEN) was not working properly if there were untried alternatives prior
1251        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1252        backtracked to try for "b" instead of moving to the next alternative branch
1253        at the same level (in this case, to look for "c"). The Perl documentation
1254        is clear that when (*THEN) is backtracked onto, it goes to the "next
1255        alternative in the innermost enclosing group".
1256    
1257    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1258        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1259        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1260        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1261        (*THEN).
1262    
1263    3.  If \s appeared in a character class, it removed the VT character from
1264        the class, even if it had been included by some previous item, for example
1265        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1266        of \s, but is part of the POSIX "space" class.)
1267    
1268    4.  A partial match never returns an empty string (because you can always
1269        match an empty string at the end of the subject); however the checking for
1270        an empty string was starting at the "start of match" point. This has been
1271        changed to the "earliest inspected character" point, because the returned
1272        data for a partial match starts at this character. This means that, for
1273        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1274        (previously it gave "no match").
1275    
1276    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1277        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1278        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1279        has an implication that the given string is incomplete (because a partial
1280        match is preferred over a full match). For this reason, these items now
1281        give a partial match in this situation. [Aside: previously, the one case
1282        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1283        match rather than a full match, which was wrong by the old rules, but is
1284        now correct.]
1285    
1286    6.  There was a bug in the handling of #-introduced comments, recognized when
1287        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1288        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1289        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1290        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1291        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1292        places in pcre_compile().
1293    
1294    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1295        comments when looking ahead for named forward references to subpatterns,
1296        the only newline sequence it recognized was NL. It now handles newlines
1297        according to the set newline convention.
1298    
1299    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1300        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1301        cater for a lack of strerror(). These oversights have been fixed.
1302    
1303    9.  Added --match-limit and --recursion-limit to pcregrep.
1304    
1305    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1306    
1307    11. When the -o option was used, pcregrep was setting a return code of 1, even
1308        when matches were found, and --line-buffered was not being honoured.
1309    
1310    12. Added an optional parentheses number to the -o and --only-matching options
1311        of pcregrep.
1312    
1313    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1314        can match an empty string. The code to do it in pcretest and pcredemo
1315        needed fixing:
1316    
1317        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1318            only one byte after an empty string match just before CRLF (this case
1319            just got forgotten; "any" and "anycrlf" were OK).
1320    
1321        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1322            mode when an empty string match preceded an ASCII character followed by
1323            a non-ASCII character. (The code for advancing by one character rather
1324            than one byte was nonsense.)
1325    
1326        (c) The pcredemo.c sample program did not have any code at all to handle
1327            the cases when CRLF is a valid newline sequence.
1328    
1329    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1330        as a starting offset was within the subject string. There is now a new
1331        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1332        negative or greater than the length of the string. In order to test this,
1333        pcretest is extended to allow the setting of negative starting offsets.
1334    
1335    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1336        starting offset points to the beginning of a UTF-8 character was
1337        unnecessarily clumsy. I tidied it up.
1338    
1339    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1340        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1341    
1342    17. Nobody had reported that the --include_dir option, which was added in
1343        release 7.7 should have been called --include-dir (hyphen, not underscore)
1344        for compatibility with GNU grep. I have changed it to --include-dir, but
1345        left --include_dir as an undocumented synonym, and the same for
1346        --exclude-dir, though that is not available in GNU grep, at least as of
1347        release 2.5.4.
1348    
1349    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1350        characters from a string of bytes) have been redefined so as not to use
1351        loops, in order to improve performance in some environments. At the same
1352        time, I abstracted some of the common code into auxiliary macros to save
1353        repetition (this should not affect the compiled code).
1354    
1355    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1356        compile-time error is now given if \c is not followed by an ASCII
1357        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1358        different, and any byte value is allowed.)
1359    
1360    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1361        START_OPTIMIZE option, which is now allowed at compile time - but just
1362        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1363        to pcregrep and other applications that have no direct access to PCRE
1364        options. The new /Y option in pcretest sets this option when calling
1365        pcre_compile().
1366    
1367    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1368        back references. Groups containing recursive back references were forced to
1369        be atomic by that change, but in the case of named groups, the amount of
1370        memory required was incorrectly computed, leading to "Failed: internal
1371        error: code overflow". This has been fixed.
1372    
1373    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1374        pcretest.c, to avoid build problems in some Borland environments.
1375    
1376    
1377    Version 8.10 25-Jun-2010
1378    ------------------------
1379    
1380    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1381        THEN.
1382    
1383    2.  (*ACCEPT) was not working when inside an atomic group.
1384    
1385    3.  Inside a character class, \B is treated as a literal by default, but
1386        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1387        causes the error). The code is unchanged, but I tidied the documentation.
1388    
1389    4.  Inside a character class, PCRE always treated \R and \X as literals,
1390        whereas Perl faults them if its -w option is set. I have changed PCRE so
1391        that it faults them when PCRE_EXTRA is set.
1392    
1393    5.  Added support for \N, which always matches any character other than
1394        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1395    
1396    6.  When compiling pcregrep with newer versions of gcc which may have
1397        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1398        declared with attribute warn_unused_result" were given. Just casting the
1399        result to (void) does not stop the warnings; a more elaborate fudge is
1400        needed. I've used a macro to implement this.
1401    
1402    7.  Minor change to pcretest.c to avoid a compiler warning.
1403    
1404    8.  Added four artifical Unicode properties to help with an option to make
1405        \s etc use properties (see next item). The new properties are: Xan
1406        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1407    
1408    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1409        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1410        this option. Modified pcretest to add /W to test this facility. Added
1411        REG_UCP to make it available via the POSIX interface.
1412    
1413    10. Added --line-buffered to pcregrep.
1414    
1415    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1416        studied, and the match started with a letter with a code point greater than
1417        127 whose first byte was different to the first byte of the other case of
1418        the letter, the other case of this starting letter was not recognized
1419        (#976).
1420    
1421    12. If a pattern that was studied started with a repeated Unicode property
1422        test, for example, \p{Nd}+, there was the theoretical possibility of
1423        setting up an incorrect bitmap of starting bytes, but fortunately it could
1424        not have actually happened in practice until change 8 above was made (it
1425        added property types that matched character-matching opcodes).
1426    
1427    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1428        possible starting bytes for non-anchored patterns.
1429    
1430    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1431        \R, and also a number of cases that involve Unicode properties, both
1432        explicit and implicit when PCRE_UCP is set.
1433    
1434    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1435        input, it could crash or give wrong results if characters with values
1436        greater than 0xc0 were present in the subject string. (Detail: it assumed
1437        UTF-8 input when processing these items.)
1438    
1439    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1440        size_t is 64-bit (#991).
1441    
1442    17. Added a check for running out of memory when PCRE is compiled with
1443        --disable-stack-for-recursion (#990).
1444    
1445    18. If the last data line in a file for pcretest does not have a newline on
1446        the end, a newline was missing in the output.
1447    
1448    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1449        less than 128) in its various bitmaps. However, there is a facility for
1450        generating tables according to the current locale when PCRE is compiled. It
1451        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1452        characters, are recognized by isspace() and therefore were getting set in
1453        these tables, and indeed these tables seem to approximate to ISO 8859. This
1454        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1455        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1456        which of course cannot start UTF-8 characters. I have changed the code so
1457        that only real ASCII characters (less than 128) and the correct starting
1458        bytes for UTF-8 encodings are set for characters greater than 127 when in
1459        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1460        altogether.)
1461    
1462    20. Added the /T option to pcretest so as to be able to run tests with non-
1463        standard character tables, thus making it possible to include the tests
1464        used for 19 above in the standard set of tests.
1465    
1466    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1467        reference to a subpattern the other side of a comment that contains an
1468        opening parenthesis caused either an internal compiling error, or a
1469        reference to the wrong subpattern.
1470    
1471    
1472    Version 8.02 19-Mar-2010
1473    ------------------------
1474    
1475    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1476    
1477    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1478        configured.
1479    
1480    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1481        original author of that file, following a query about its status.
1482    
1483    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1484        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1485    
1486    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1487        quantifier applied to a forward-referencing subroutine call, could compile
1488        incorrect code or give the error "internal error: previously-checked
1489        referenced subpattern not found".
1490    
1491    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1492        variables to point to external functions. For these systems, therefore,
1493        pcre_malloc etc. are now initialized to local functions that call the
1494        relevant global functions.
1495    
1496    7.  There were two entries missing in the vectors called coptable and poptable
1497        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1498        I've fixed the data, and added a kludgy way of testing at compile time that
1499        the lengths are correct (equal to the number of opcodes).
1500    
1501    8.  Following on from 7, I added a similar kludge to check the length of the
1502        eint vector in pcreposix.c.
1503    
1504    9.  Error texts for pcre_compile() are held as one long string to avoid too
1505        much relocation at load time. To find a text, the string is searched,
1506        counting zeros. There was no check for running off the end of the string,
1507        which could happen if a new error number was added without updating the
1508        string.
1509    
1510    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1511    
1512    11. \K was not working if it appeared in an atomic group or in a group that
1513        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1514        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1515        the assertion is positive, but not if it is negative.
1516    
1517    12. Change 11 fortuitously reduced the size of the stack frame used in the
1518        "match()" function of pcre_exec.c by one pointer. Forthcoming
1519        implementation of support for (*MARK) will need an extra pointer on the
1520        stack; I have reserved it now, so that the stack frame size does not
1521        decrease.
1522    
1523    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1524        item in branch that calls a recursion is a subroutine call - as in the
1525        second branch in the above example - was incorrectly given the compile-
1526        time error "recursive call could loop indefinitely" because pcre_compile()
1527        was not correctly checking the subroutine for matching a non-empty string.
1528    
1529    14. The checks for overrunning compiling workspace could trigger after an
1530        overrun had occurred. This is a "should never occur" error, but it can be
1531        triggered by pathological patterns such as hundreds of nested parentheses.
1532        The checks now trigger 100 bytes before the end of the workspace.
1533    
1534    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1535    
1536    
1537    Version 8.01 19-Jan-2010
1538    ------------------------
1539    
1540    1.  If a pattern contained a conditional subpattern with only one branch (in
1541        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1542        computed the wrong minimum data length (which is of course zero for such
1543        subpatterns). This could cause incorrect "no match" results.
1544    
1545    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1546        the pattern is reset in the first branch, pcre_compile() failed with
1547        "internal error: code overflow at offset...". This happened only when
1548        the reset was to the original external option setting. (An optimization
1549        abstracts leading options settings into an external setting, which was the
1550        cause of this.)
1551    
1552    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1553        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1554        assertion pattern did not match (meaning that the assertion was true), it
1555        was incorrectly treated as false if the SKIP had been reached during the
1556        matching. This also applied to assertions used as conditions.
1557    
1558    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1559        assertion subpattern, including such a pattern used as a condition,
1560        unpredictable results occurred, instead of the error return
1561        PCRE_ERROR_DFA_UITEM.
1562    
1563    5.  The C++ GlobalReplace function was not working like Perl for the special
1564        situation when an empty string is matched. It now does the fancy magic
1565        stuff that is necessary.
1566    
1567    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1568        removed. (These were left over from very, very early versions of PCRE.)
1569    
1570    7.  Some cosmetic changes to the code to make life easier when compiling it
1571        as part of something else:
1572    
1573        (a) Change DEBUG to PCRE_DEBUG.
1574    
1575        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1576            called "current" as "current_branch", to prevent a collision with the
1577            Linux macro when compiled as a kernel module.
1578    
1579        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1580            prevent a collision with the Linux macro when compiled as a kernel
1581            module.
1582    
1583    8.  In pcre_compile() there are some checks for integer overflows that used to
1584        cast potentially large values to (double). This has been changed to that
1585        when building, a check for int64_t is made, and if it is found, it is used
1586        instead, thus avoiding the use of floating point arithmetic. (There is no
1587        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1588        double.
1589    
1590    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1591        2005 (difference between two addresses compared to an unsigned value).
1592    
1593    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1594        custom one, because of the following reported problem in Windows:
1595    
1596          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1597              under Win32.
1598          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1599              therefore missing the function definition.
1600          - The compiler thus generates a "C" signature for the test function.
1601          - The linker fails to find the "C" function.
1602          - PCRE fails to configure if asked to do so against libbz2.
1603    
1604    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1605        messages were output:
1606    
1607          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1608          rerunning libtoolize, to keep the correct libtool macros in-tree.
1609          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1610    
1611        I have done both of these things.
1612    
1613    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1614        most of the time, it *can* run out if it is given a pattern that contains a
1615        runaway infinite recursion. I updated the discussion in the pcrestack man
1616        page.
1617    
1618    13. Now that we have gone to the x.xx style of version numbers, the minor
1619        version may start with zero. Using 08 or 09 is a bad idea because users
1620        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1621        interpreted as invalid octal numbers. I've updated the previous comment in
1622        configure.ac, and also added a check that gives an error if 08 or 09 are
1623        used.
1624    
1625    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1626        causing partial matching to fail when the end of the subject matched \W
1627        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1628    
1629    15. There were some discrepancies between the declarations in pcre_internal.h
1630        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1631        their definitions. The declarations used "const uschar *" and the
1632        definitions used USPTR. Even though USPTR is normally defined as "const
1633        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1634        reported that: "This difference in casting confuses some C++ compilers, for
1635        example, SunCC recognizes above declarations as different functions and
1636        generates broken code for hbpcre." I have changed the declarations to use
1637        USPTR.
1638    
1639    16. GNU libtool is named differently on some systems. The autogen.sh script now
1640        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1641        (FreeBSD).
1642    
1643    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1644        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1645        comment: "Figure out how to create a longlong from a string: strtoll and
1646        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1647        instance, but it only takes 2 args instead of 3!"
1648    
1649    18. A subtle bug concerned with back references has been fixed by a change of
1650        specification, with a corresponding code fix. A pattern such as
1651        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1652        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1653        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1654        same bug. Such groups have to be quantified to be useful, or contained
1655        inside another quantified group. (If there's no repetition, the reference
1656        can never match.) The problem arises because, having left the group and
1657        moved on to the rest of the pattern, a later failure that backtracks into
1658        the group uses the captured value from the final iteration of the group
1659        rather than the correct earlier one. I have fixed this in PCRE by forcing
1660        any group that contains a reference to itself to be an atomic group; that
1661        is, there cannot be any backtracking into it once it has completed. This is
1662        similar to recursive and subroutine calls.
1663    
1664    
1665    Version 8.00 19-Oct-09
1666    ----------------------
1667    
1668    1.  The table for translating pcre_compile() error codes into POSIX error codes
1669        was out-of-date, and there was no check on the pcre_compile() error code
1670        being within the table. This could lead to an OK return being given in
1671        error.
1672    
1673    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1674        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1675        in a Windows environment.
1676    
1677    3.  The pcregrep --count option prints the count for each file even when it is
1678        zero, as does GNU grep. However, pcregrep was also printing all files when
1679        --files-with-matches was added. Now, when both options are given, it prints
1680        counts only for those files that have at least one match. (GNU grep just
1681        prints the file name in this circumstance, but including the count seems
1682        more useful - otherwise, why use --count?) Also ensured that the
1683        combination -clh just lists non-zero counts, with no names.
1684    
1685    4.  The long form of the pcregrep -F option was incorrectly implemented as
1686        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1687        but it seems right to fix it, and I didn't think it was worth preserving
1688        the old behaviour.
1689    
1690    5.  The command line items --regex=pattern and --regexp=pattern were not
1691        recognized by pcregrep, which required --regex pattern or --regexp pattern
1692        (with a space rather than an '='). The man page documented the '=' forms,
1693        which are compatible with GNU grep; these now work.
1694    
1695    6.  No libpcreposix.pc file was created for pkg-config; there was just
1696        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1697    
1698    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1699        when UCP support is not needed, by modifying the Python script that
1700        generates it from Unicode data files. This should not matter if the module
1701        is correctly used as a library, but I received one complaint about 50K of
1702        unwanted data. My guess is that the person linked everything into his
1703        program rather than using a library. Anyway, it does no harm.
1704    
1705    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1706        was a minimum greater than 1 for a wide character in a possessive
1707        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1708        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1709        character. Chaos in the form of incorrect output or a compiling loop could
1710        result.
1711    
1712    9.  The restrictions on what a pattern can contain when partial matching is
1713        requested for pcre_exec() have been removed. All patterns can now be
1714        partially matched by this function. In addition, if there are at least two
1715        slots in the offset vector, the offset of the earliest inspected character
1716        for the match and the offset of the end of the subject are set in them when
1717        PCRE_ERROR_PARTIAL is returned.
1718    
1719    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1720        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1721        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1722        and may be more useful for multi-segment matching.
1723    
1724    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1725        used to be given if ever the end of the subject was reached; now it is
1726        given only if matching could not proceed because another character was
1727        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1728        string "Z", which now yields "no match" instead of "partial match". In the
1729        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1730        final character ended with (*FAIL).
1731    
1732    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1733        if the pattern had a "must contain" character that was already found in the
1734        earlier partial match, unless partial matching was again requested. For
1735        example, with the pattern /dog.(body)?/, the "must contain" character is
1736        "g". If the first part-match was for the string "dog", restarting with
1737        "sbody" failed. This bug has been fixed.
1738    
1739    13. The string returned by pcre_dfa_exec() after a partial match has been
1740        changed so that it starts at the first inspected character rather than the
1741        first character of the match. This makes a difference only if the pattern
1742        starts with a lookbehind assertion or \b or \B (\K is not supported by
1743        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1744        matching functions compatible, and I think it's the right thing to do.
1745    
1746    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1747        so that the demonstration program is easily available in environments where
1748        PCRE has not been installed from source.
1749    
1750    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1751        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1752        library.
1753    
1754    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1755        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1756        is not the first non-POSIX option to be added. Clearly some people find
1757        these options useful.
1758    
1759    17. If a caller to the POSIX matching function regexec() passes a non-zero
1760        value for nmatch with a NULL value for pmatch, the value of
1761        nmatch is forced to zero.
1762    
1763    18. RunGrepTest did not have a test for the availability of the -u option of
1764        the diff command, as RunTest does. It now checks in the same way as
1765        RunTest, and also checks for the -b option.
1766    
1767    19. If an odd number of negated classes containing just a single character
1768        interposed, within parentheses, between a forward reference to a named
1769        subpattern and the definition of the subpattern, compilation crashed with
1770        an internal error, complaining that it could not find the referenced
1771        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1772        [The bug was that it was starting one character too far in when skipping
1773        over the character class, thus treating the ] as data rather than
1774        terminating the class. This meant it could skip too much.]
1775    
1776    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1777        /g option in pcretest when the pattern contains \K, which makes it possible
1778        to have an empty string match not at the start, even when the pattern is
1779        anchored. Updated pcretest and pcredemo to use this option.
1780    
1781    21. If the maximum number of capturing subpatterns in a recursion was greater
1782        than the maximum at the outer level, the higher number was returned, but
1783        with unset values at the outer level. The correct (outer level) value is
1784        now given.
1785    
1786    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1787        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1788        make it do so. The string so far is captured, making this feature
1789        compatible with Perl.
1790    
1791    23. The tests have been re-organized, adding tests 11 and 12, to make it
1792        possible to check the Perl 5.10 features against Perl 5.10.
1793    
1794    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1795        pattern matches a fixed length string. PCRE did not allow this; now it
1796        does. Neither allows recursion.
1797    
1798    25. I finally figured out how to implement a request to provide the minimum
1799        length of subject string that was needed in order to match a given pattern.
1800        (It was back references and recursion that I had previously got hung up
1801        on.) This code has now been added to pcre_study(); it finds a lower bound
1802        to the length of subject needed. It is not necessarily the greatest lower
1803        bound, but using it to avoid searching strings that are too short does give
1804        some useful speed-ups. The value is available to calling programs via
1805        pcre_fullinfo().
1806    
1807    26. While implementing 25, I discovered to my embarrassment that pcretest had
1808        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1809        study optimizations had never been tested with that matching function.
1810        Oops. What is worse, even when it was passed study data, there was a bug in
1811        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1812        were also very few tests of studied patterns with pcre_dfa_exec().
1813    
1814    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1815        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1816        on the other side of the coin, they are no longer allowed to have different
1817        names, because these cannot be distinguished in PCRE, and this has caused
1818        confusion. (This is a difference from Perl.)
1819    
1820    28. When duplicate subpattern names are present (necessarily with different
1821        numbers, as required by 27 above), and a test is made by name in a
1822        conditional pattern, either for a subpattern having been matched, or for
1823        recursion in such a pattern, all the associated numbered subpatterns are
1824        tested, and the overall condition is true if the condition is true for any
1825        one of them. This is the way Perl works, and is also more like the way
1826        testing by number works.
1827    
1828    
1829    Version 7.9 11-Apr-09
1830    ---------------------
1831    
1832    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1833        (pcretest), all targets were linked against these libraries. This included
1834        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1835        libraries. This caused unwanted dependencies to be created. This problem
1836        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1837        pcretest is linked with readline.
1838    
1839    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1840        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1841        moved outside it again, because FALSE and TRUE are already defined in AIX,
1842        but BOOL is not.
1843    
1844    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1845        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1846    
1847    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1848        hyphens) following file names and line numbers when outputting matching
1849        lines. This is not true; no spaces are inserted. I have also clarified the
1850        wording for the --colour (or --color) option.
1851    
1852    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1853        was not coloured; this is different to GNU grep, so I have changed it to be
1854        the same.
1855    
1856    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1857        each matching line was coloured. Now it goes on to look for further matches
1858        of any of the test patterns, which is the same behaviour as GNU grep.
1859    
1860    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1861        doesn't make sense to accept an empty string match in pcregrep, so I have
1862        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1863        seems to be how GNU grep behaves. [But see later change 40 for release
1864        8.33.]
1865    
1866    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1867        start or after a newline", because the conditional assertion was not being
1868        correctly handled. The rule now is that both the assertion and what follows
1869        in the first alternative must satisfy the test.
1870    
1871    9.  If auto-callout was enabled in a pattern with a conditional group whose
1872        condition was an assertion, PCRE could crash during matching, both with
1873        pcre_exec() and pcre_dfa_exec().
1874    
1875    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1876        used for matching.
1877    
1878    11. Unicode property support in character classes was not working for
1879        characters (bytes) greater than 127 when not in UTF-8 mode.
1880    
1881    12. Added the -M command line option to pcretest.
1882    
1883    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1884    
1885    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1886    
1887    16. Added comments and documentation about mis-use of no_arg in the C++
1888        wrapper.
1889    
1890    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1891        from Martin Jerabek that uses macro names for all relevant character and
1892        string constants.
1893    
1894    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1895        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1896        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1897        these, but not everybody uses configure.
1898    
1899    19. A conditional group that had only one branch was not being correctly
1900        recognized as an item that could match an empty string. This meant that an
1901        enclosing group might also not be so recognized, causing infinite looping
1902        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1903        with the subject "ab", where knowledge that the repeated group can match
1904        nothing is needed in order to break the loop.
1905    
1906    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1907        exec(), but without supplying a callout function, matching went wrong.
1908    
1909    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1910        leak if the size of the offset vector was greater than 30. When the vector
1911        is smaller, the saved offsets during recursion go onto a local stack
1912        vector, but for larger vectors malloc() is used. It was failing to free
1913        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1914        error, in fact).
1915    
1916    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1917        heapframe that is used only when UTF-8 support is enabled. This caused no
1918        problem, but was untidy.
1919    
1920    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1921        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1922        included within another project.
1923    
1924    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1925        slightly modified by me:
1926    
1927          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1928              not building pcregrep.
1929    
1930          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1931              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1932    
1933    25. Forward references, both numeric and by name, in patterns that made use of
1934        duplicate group numbers, could behave incorrectly or give incorrect errors,
1935        because when scanning forward to find the reference group, PCRE was not
1936        taking into account the duplicate group numbers. A pattern such as
1937        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1938    
1939    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1940        the feature of a custom pointer more persuasive (as requested by a user).
1941    
1942    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1943        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1944        pre-defined.
1945    
1946    28. Added support for (*UTF8) at the start of a pattern.
1947    
1948    29. Arrange for flags added by the "release type" setting in CMake to be shown
1949        in the configuration summary.
1950    
1951    
1952    Version 7.8 05-Sep-08
1953    ---------------------
1954    
1955    1.  Replaced UCP searching code with optimized version as implemented for Ad
1956        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1957        stage table and inline lookup instead of a function, giving speed ups of 2
1958        to 5 times on some simple patterns that I tested. Permission was given to
1959        distribute the MultiStage2.py script that generates the tables (it's not in
1960        the tarball, but is in the Subversion repository).
1961    
1962    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1963        scripts.
1964    
1965    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1966        a group with a zero qualifier. The result of the study could be incorrect,
1967        or the function might crash, depending on the pattern.
1968    
1969    4.  Caseless matching was not working for non-ASCII characters in back
1970        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1971        It now works when Unicode Property Support is available.
1972    
1973    5.  In pcretest, an escape such as \x{de} in the data was always generating
1974        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1975        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1976        truncation.
1977    
1978    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1979    
1980    7.  Added two (int) casts to pcregrep when printing the difference of two
1981        pointers, in case they are 64-bit values.
1982    
1983    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1984        test 2 if it fails.
1985    
1986    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
1987        and a #define of that name to empty if it is not externally set. This is to
1988        allow users of MSVC to set it if necessary.
1989    
1990    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
1991        the convenience functions in the pcre_get.c source file.
1992    
1993    11. An option change at the start of a pattern that had top-level alternatives
1994        could cause overwriting and/or a crash. This command provoked a crash in
1995        some environments:
1996    
1997          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
1998    
1999        This potential security problem was recorded as CVE-2008-2371.
2000    
2001    12. For a pattern where the match had to start at the beginning or immediately
2002        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2003        pcre_dfa_exec() could read past the end of the passed subject if there was
2004        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2005        pcretest so that it places the subject at the end of its malloc-ed buffer.
2006    
2007    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2008        exec() might read past the end of the data buffer in UTF-8 mode.
2009    
2010    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2011        the data contained the byte 0x85 as part of a UTF-8 character within its
2012        first line. This applied both to normal and DFA matching.
2013    
2014    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2015        /^[^d]*?$/8 failed to match "abc".
2016    
2017    16. Added a missing copyright notice to pcrecpp_internal.h.
2018    
2019    17. Make it more clear in the documentation that values returned from
2020        pcre_exec() in ovector are byte offsets, not character counts.
2021    
2022    18. Tidied a few places to stop certain compilers from issuing warnings.
2023    
2024    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2025        supplied by Stefan Weber. I made a further small update for 7.8 because
2026        there is a change of source arrangements: the pcre_searchfuncs.c module is
2027        replaced by pcre_ucd.c.
2028    
2029    
2030    Version 7.7 07-May-08
2031    ---------------------
2032    
2033    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2034        a string to a long long, pretend we don't even have a long long." This is
2035        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2036    
2037    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2038        pre-7.6 versions, which defined a global no_arg variable instead of putting
2039        it in the RE class. (See also #8 below.)
2040    
2041    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2042        Lopes.
2043    
2044    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2045    
2046        (1) The include/exclude patterns were being applied to the whole pathnames
2047            of files, instead of just to the final components.
2048    
2049        (2) If there was more than one level of directory, the subdirectories were
2050            skipped unless they satisfied the include/exclude conditions. This is
2051            inconsistent with GNU grep (and could even be seen as contrary to the
2052            pcregrep specification - which I improved to make it absolutely clear).
2053            The action now is always to scan all levels of directory, and just
2054            apply the include/exclude patterns to regular files.
2055    
2056    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2057        --exclude_dir in the tests to avoid scanning .svn directories.
2058    
2059    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2060        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2061        doesn't support NULs in patterns.
2062    
2063    7.  Added some missing "const"s to declarations of static tables in
2064        pcre_compile.c and pcre_dfa_exec.c.
2065    
2066    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2067        caused by fix #2  above. (Subsequently also a second patch to fix the
2068        first patch. And a third patch - this was a messy problem.)
2069    
2070    9.  Applied Craig's patch to remove the use of push_back().
2071    
2072    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2073        matching function regexec().
2074    
2075    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2076        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2077        references. PCRE supports relative numbers with this syntax (I don't think
2078        Oniguruma does).
2079    
2080    12. Previously, a group with a zero repeat such as (...){0} was completely
2081        omitted from the compiled regex. However, this means that if the group
2082        was called as a subroutine from elsewhere in the pattern, things went wrong
2083        (an internal error was given). Such groups are now left in the compiled
2084        pattern, with a new opcode that causes them to be skipped at execution
2085        time.
2086    
2087    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2088        to the way PCRE behaves:
2089    
2090        (a) A lone ] character is dis-allowed (Perl treats it as data).
2091    
2092        (b) A back reference to an unmatched subpattern matches an empty string
2093            (Perl fails the current match path).
2094    
2095        (c) A data ] in a character class must be notated as \] because if the
2096            first data character in a class is ], it defines an empty class. (In
2097            Perl it is not possible to have an empty class.) The empty class []
2098            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2099            The negative empty class [^] matches any one character, independently
2100            of the DOTALL setting.
2101    
2102    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2103        non-existent subpattern following a character class starting with ']' and
2104        containing () gave an internal compiling error instead of "reference to
2105        non-existent subpattern". Fortunately, when the pattern did exist, the
2106        compiled code was correct. (When scanning forwards to check for the
2107        existence of the subpattern, it was treating the data ']' as terminating
2108        the class, so got the count wrong. When actually compiling, the reference
2109        was subsequently set up correctly.)
2110    
2111    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2112        it was being rejected as not supported by pcre_dfa_exec(), even though
2113        other assertions are supported. I have made pcre_dfa_exec() support
2114        (*FAIL).
2115    
2116    16. The implementation of 13c above involved the invention of a new opcode,
2117        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2118        cannot be changed at match time, I realized I could make a small
2119        improvement to matching performance by compiling OP_ALLANY instead of
2120        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2121        on the OP_ANY path.
2122    
2123    17. Compiling pcretest on Windows with readline support failed without the
2124        following two fixes: (1) Make the unistd.h include conditional on
2125        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2126    
2127    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2128        ncurses library to be included for pcretest when ReadLine support is
2129        requested, but also to allow for it to be overridden. This patch came from
2130        Daniel Bergström.
2131    
2132    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2133        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2134        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2135        spotting this.
2136    
2137    
2138    Version 7.6 28-Jan-08
2139    ---------------------
2140    
2141    1.  A character class containing a very large number of characters with
2142        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2143        overflow.
2144    
2145    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2146        HAVE_LONG_LONG is not defined.
2147    
2148    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2149        bring them up to date and include new features. This patch includes:
2150    
2151        - Fixed PH's badly added libz and libbz2 support.
2152        - Fixed a problem with static linking.
2153        - Added pcredemo. [But later removed - see 7 below.]
2154        - Fixed dftables problem and added an option.
2155        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2156            HAVE_LONG_LONG.
2157        - Added readline support for pcretest.
2158        - Added an listing of the option settings after cmake has run.
2159    
2160    4.  A user submitted a patch to Makefile that makes it easy to create
2161        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2162        Makefile.am that cause it to include this special target, without
2163        affecting anything else. Note that the same mingw target plus all
2164        the other distribution libraries and programs are now supported
2165        when configuring with CMake (see 6 below) instead of with
2166        Configure/Make.
2167    
2168    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2169        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2170        exported in the Windows port". It has not yet been confirmed that the patch
2171        solves the problem, but it does no harm.
2172    
2173    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2174        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2175        with CMake, and also correct the comment about stack recursion.
2176    
2177    7.  Remove the automatic building of pcredemo from the ./configure system and
2178        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2179        of a program that users should build themselves after PCRE is installed, so
2180        building it automatically is not really right. What is more, it gave
2181        trouble in some build environments.
2182    
2183    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2184    
2185    
2186    Version 7.5 10-Jan-08
2187    ---------------------
2188    
2189    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2190        values in parens when parsing an RE using the C++ wrapper."
2191    
2192    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2193        Characters greater than 255 were excluded from the class instead of being
2194        included.
2195    
2196    3.  The same bug as (2) above applied to negated POSIX classes such as
2197        [:^space:].
2198    
2199    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2200        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2201        I have changed it.
2202    
2203    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2204        first named subpattern) and a construct such as (?&a) would reference the
2205        first named subpattern whose name started with "a" (in other words, the
2206        length check was missing). Both these problems are fixed. "Subpattern name
2207        expected" is now given for (?&) (a zero-length name), and this patch also
2208        makes it give the same error for \k'' (previously it complained that that
2209        was a reference to a non-existent subpattern).
2210    
2211    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2212        this is right because (?- can be followed by option settings as well as by
2213        digits. I have, however, made the messages clearer.
2214    
2215    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2216        than the number used in the conditional) now cause a compile-time error.
2217        This is actually not compatible with Perl, which accepts such patterns, but
2218        treats the conditional as always being FALSE (as PCRE used to), but it
2219        seems to me that giving a diagnostic is better.
2220    
2221    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2222        and messages.
2223    
2224    9.  Fix two occurrences of "backslash" in comments that should have been
2225        "backspace".
2226    
2227    10. Remove two redundant lines of code that can never be obeyed (their function
2228        was moved elsewhere).
2229    
2230    11. The program that makes PCRE's Unicode character property table had a bug
2231        which caused it to generate incorrect table entries for sequences of
2232        characters that have the same character type, but are in different scripts.
2233        It amalgamated them into a single range, with the script of the first of
2234        them. In other words, some characters were in the wrong script. There were
2235        thirteen such cases, affecting characters in the following ranges:
2236    
2237          U+002b0 - U+002c1
2238          U+0060c - U+0060d
2239          U+0061e - U+00612
2240          U+0064b - U+0065e
2241          U+0074d - U+0076d
2242          U+01800 - U+01805
2243          U+01d00 - U+01d77
2244          U+01d9b - U+01dbf
2245          U+0200b - U+0200f
2246          U+030fc - U+030fe
2247          U+03260 - U+0327f
2248          U+0fb46 - U+0fbb1
2249          U+10450 - U+1049d
2250    
2251    12. The -o option (show only the matching part of a line) for pcregrep was not
2252        compatible with GNU grep in that, if there was more than one match in a
2253        line, it showed only the first of them. It now behaves in the same way as
2254        GNU grep.
2255    
2256    13. If the -o and -v options were combined for pcregrep, it printed a blank
2257        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2258        does the same. The return code can be used to tell if there were any
2259        non-matching lines.
2260    
2261    14. Added --file-offsets and --line-offsets to pcregrep.
2262    
2263    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2264        infinitely looping recursion. The bug was that positive lookaheads were not
2265        being skipped when checking for a possible empty match (negative lookaheads
2266        and both kinds of lookbehind were skipped).
2267    
2268    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2269        inclusion of <windows.h> to before rather than after the definition of
2270        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2271    
2272    17. Specifying a possessive quantifier with a specific limit for a Unicode
2273        character property caused pcre_compile() to compile bad code, which led at
2274        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2275        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2276        caused the error; without that there was no problem.
2277    
2278    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2279    
2280    19. Added --enable-pcretest-libreadline.
2281    
2282    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2283        RE::GlobalReplace(). As a result, the number of replacements returned was
2284        double what it should be. I removed one of the increments, but Craig sent a
2285        later patch that removed the other one (the right fix) and added unit tests
2286        that check the return values (which was not done before).
2287    
2288    21. Several CMake things:
2289    
2290        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2291            the names libpcre and libpcreposix, not just pcre and pcreposix.
2292    
2293        (2) The above change means that pcretest and pcregrep are now correctly
2294            linked with the newly-built libraries, not previously installed ones.
2295    
2296        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2297    
2298    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2299        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2300        UTF-8 newline character). The key issue is that the pattern starts .*;
2301        this means that the match must be either at the beginning, or after a
2302        newline. The bug was in the code for advancing after a failed match and
2303        checking that the new position followed a newline. It was not taking
2304        account of UTF-8 characters correctly.
2305    
2306    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2307        character classes. PCRE was not treating the sequence [:...:] as a
2308        character class unless the ... were all letters. Perl, however, seems to
2309        allow any characters between [: and :], though of course it rejects as
2310        unknown any "names" that contain non-letters, because all the known class
2311        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2312        for example, whereas PCRE did not - it did not recognize a POSIX character
2313        class. This seemed a bit dangerous, so the code has been changed to be
2314        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2315        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2316        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2317        Perl does, and where it didn't before.
2318    
2319    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2320        Windows environments %n is disabled by default.
2321    
2322    
2323    Version 7.4 21-Sep-07
2324    ---------------------
2325    
2326    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2327        means that a class such as [\s] counted as "explicit reference to CR or
2328        LF". That isn't really right - the whole point of the change was to try to
2329        help when there was an actual mention of one of the two characters. So now
2330        the change happens only if \r or \n (or a literal CR or LF) character is
2331        encountered.
2332    
2333    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2334        of both had grown to the point where there were only 3 bits left.
2335        Fortunately, there was spare space in the data structure, and so I have
2336        moved the internal flags into a new 16-bit field to free up more option
2337        bits.
2338    
2339    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2340        but did not set the internal JCHANGED flag - either of these is enough to
2341        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2342        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2343        start sets both bits.
2344    
2345    4.  Added options (at build time, compile time, exec time) to change \R from
2346        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2347    
2348    5.  doc/pcresyntax.html was missing from the distribution.
2349    
2350    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2351        compatibility, even though it is no longer used.
2352    
2353    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2354        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2355        windows.h file is present (where different names are used). [This was
2356        reversed later after testing - see 16 below.]
2357    
2358    8.  Changed all #include <config.h> to #include "config.h". There were also
2359        some further <pcre.h> cases that I changed to "pcre.h".
2360    
2361    9.  When pcregrep was used with the --colour option, it missed the line ending
2362        sequence off the lines that it output.
2363    
2364    10. It was pointed out to me that arrays of string pointers cause lots of
2365        relocations when a shared library is dynamically loaded. A technique of
2366        using a single long string with a table of offsets can drastically reduce
2367        these. I have refactored PCRE in four places to do this. The result is
2368        dramatic:
2369    
2370          Originally:                          290
2371          After changing UCP table:            187
2372          After changing error message table:   43
2373          After changing table of "verbs"       36
2374          After changing table of Posix names   22
2375    
2376        Thanks to the folks working on Gregex for glib for this insight.
2377    
2378    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2379        unicode-properties was also set.
2380    
2381    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2382    
2383    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2384        checked only for CRLF.
2385    
2386    14. Added casts to pcretest.c to avoid compiler warnings.
2387    
2388    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2389    
2390    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2391        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2392        entirely. This removes changes made in 7 above.
2393    
2394    17. The CMake files have been updated, and there is now more information about
2395        building with CMake in the NON-UNIX-USE document.
2396    
2397    
2398    Version 7.3 28-Aug-07
2399  ---------------------  ---------------------
2400    
2401   1. In the rejigging of the build system that eventually resulted in 7.1, the   1. In the rejigging of the build system that eventually resulted in 7.1, the
# Line 64  Version 7.3 05-Jul-07 Line 2458  Version 7.3 05-Jul-07
2458      dynamic way, which I have now done. The artificial limitation on group      dynamic way, which I have now done. The artificial limitation on group
2459      length has been removed - we now have only the limit on the total length of      length has been removed - we now have only the limit on the total length of
2460      the compiled pattern, which depends on the LINK_SIZE setting.      the compiled pattern, which depends on the LINK_SIZE setting.
2461    
2462  10. Fixed a bug in the documentation for get/copy named substring when  10. Fixed a bug in the documentation for get/copy named substring when
2463      duplicate names are permitted. If none of the named substrings are set, the      duplicate names are permitted. If none of the named substrings are set, the
2464      functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an      functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2465      empty string.      empty string.
2466    
2467  11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E  11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2468      instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,      instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2469      because the ] is interpreted as the first data character and the      because the ] is interpreted as the first data character and the
2470      terminating ] is not found. PCRE has been made compatible with Perl in this      terminating ] is not found. PCRE has been made compatible with Perl in this
2471      regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could      regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2472      cause memory overwriting.      cause memory overwriting.
2473    
2474  10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty  10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2475      string has been matched (to stop an infinite loop). It was not recognizing      string has been matched (to stop an infinite loop). It was not recognizing
2476      a conditional subpattern that could match an empty string if that      a conditional subpattern that could match an empty string if that
2477      subpattern was within another subpattern. For example, it looped when      subpattern was within another subpattern. For example, it looped when
2478      trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the      trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2479      condition was not nested. This bug has been fixed.      condition was not nested. This bug has been fixed.
2480    
2481  12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack  12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2482      past the start of the subject in the presence of bytes with the top bit      past the start of the subject in the presence of bytes with the top bit
2483      set, for example "\x8aBCD".      set, for example "\x8aBCD".
2484    
2485    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2486        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2487    
2488    14. Optimized (?!) to (*FAIL).
2489    
2490    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2491        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2492        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2493        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2494        does: it's just the validity check that is more restrictive.
2495    
2496    16. Inserted checks for integer overflows during escape sequence (backslash)
2497        processing, and also fixed erroneous offset values for syntax errors during
2498        backslash processing.
2499    
2500    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2501        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2502    
2503    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2504        caused an overrun.
2505    
2506    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2507        something other than just ASCII characters) inside a group that had an
2508        unlimited repeat caused a loop at compile time (while checking to see
2509        whether the group could match an empty string).
2510    
2511    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2512        [\P{Any}] did so. (Error in the code for printing property names.)
2513    
2514    21. An orphan \E inside a character class could cause a crash.
2515    
2516    22. A repeated capturing bracket such as (A)? could cause a wild memory
2517        reference during compilation.
2518    
2519    23. There are several functions in pcre_compile() that scan along a compiled
2520        expression for various reasons (e.g. to see if it's fixed length for look
2521        behind). There were bugs in these functions when a repeated \p or \P was
2522        present in the pattern. These operators have additional parameters compared
2523        with \d, etc, and these were not being taken into account when moving along
2524        the compiled data. Specifically:
2525    
2526        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2527            length.
2528    
2529        (b) An item such as \pL+ within a repeated group could cause crashes or
2530            loops.
2531    
2532        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2533            "reference to non-existent subpattern" error.
2534    
2535        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2536    
2537    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2538        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2539    
2540    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2541    
2542    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2543        character were causing crashes (broken optimization).
2544    
2545    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2546        \p or \P) caused a compile-time loop.
2547    
2548    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2549        break. For example, the unstudied pattern [\r\n]A does not match the string
2550        "\r\nA" because change 7.0/46 below moves the current point on by two
2551        characters after failing to match at the start. However, the pattern \nA
2552        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2553        the same is true. There doesn't seem any very clean way out of this, but
2554        what I have chosen to do makes the common cases work: PCRE now takes note
2555        of whether there can be an explicit match for \r or \n anywhere in the
2556        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2557        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2558        pattern has explicit CR or LF references.
2559    
2560    29. Added (*CR) etc for changing newline setting at start of pattern.
2561    
2562    
2563  Version 7.2 19-Jun-07  Version 7.2 19-Jun-07
2564  ---------------------  ---------------------

Legend:
Removed from v.207  
changed lines
  Added in v.1400

  ViewVC Help
Powered by ViewVC 1.1.5