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

Diff of /code/trunk/ChangeLog

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

revision 167 by ph10, Wed May 9 15:53:54 2007 UTC revision 1407 by ph10, Sun Dec 1 16:31:40 2013 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.2 01-May-07  Version 8.34 19-November-2013
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 default ("C" locale) set of
93        characters that match \s and are generally treated as white space,
94        following this same change in Perl 5.18. There is now no difference between
95        "Perl space" and "POSIX space". Whether VT is treated as white space in
96        other locales depends on the locale.
97    
98    19. The code for checking named groups as conditions, either for being set or
99        for being recursed, has been refactored (this is related to 14 and 15
100        above). Processing unduplicated named groups should now be as fast at
101        numerical groups, and processing duplicated groups should be faster than
102        before.
103    
104    20. Two patches to the CMake build system, by Alexander Barkov:
105    
106          (1) Replace the "source" command by "." in CMakeLists.txt because
107              "source" is a bash-ism.
108    
109          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
110              without these the CMake build does not work on Solaris.
111    
112    21. Perl has changed its handling of \8 and \9. If there is no previously
113        encountered capturing group of those numbers, they are treated as the
114        literal characters 8 and 9 instead of a binary zero followed by the
115        literals. PCRE now does the same.
116    
117    22. Following Perl, added \o{} to specify codepoints in octal, making it
118        possible to specify values greater than 0777 and also making them
119        unambiguous.
120    
121    23. Perl now gives an error for missing closing braces after \x{... instead of
122        treating the string as literal. PCRE now does the same.
123    
124    24. RunTest used to grumble if an inappropriate test was selected explicitly,
125        but just skip it when running all tests. This make it awkward to run ranges
126        of tests when one of them was inappropriate. Now it just skips any
127        inappropriate tests, as it always did when running all tests.
128    
129    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
130        character types such as \d or \w, too many callouts were inserted, and the
131        data that they returned was rubbish.
132    
133    26. In UCP mode, \s was not matching two of the characters that Perl matches,
134        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
135        were matched by \h. The code has now been refactored so that the lists of
136        the horizontal and vertical whitespace characters used for \h and \v (which
137        are defined only in one place) are now also used for \s.
138    
139    27. Add JIT support for the 64 bit TileGX architecture.
140        Patch by Jiong Wang (Tilera Corporation).
141    
142    28. Possessive quantifiers for classes (both explicit and automatically
143        generated) now use special opcodes instead of wrapping in ONCE brackets.
144    
145    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
146        (because it's meaningless), this was not happening when PCRE_CASELESS was
147        set. Not wrong, but inefficient.
148    
149    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
150        properties for \w, \d, etc) is present in a test regex. Otherwise if the
151        test contains no characters greater than 255, Perl doesn't realise it
152        should be using Unicode semantics.
153    
154    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
155        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
156        does in Unicode mode.
157    
158    32. Added the "forbid" facility to pcretest so that putting tests into the
159        wrong test files can sometimes be quickly detected.
160    
161    33. There is now a limit (default 250) on the depth of nesting of parentheses.
162        This limit is imposed to control the amount of system stack used at compile
163        time. It can be changed at build time by --with-parens-nest-limit=xxx or
164        the equivalent in CMake.
165    
166    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
167        errors. Perl warns for these when in warning mode, but PCRE has no facility
168        for giving warnings.
169    
170    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
171        However, this was not working for (?!) because it is optimized to (*FAIL),
172        for which PCRE does not allow quantifiers. The optimization is now disabled
173        when a quantifier follows (?!). I can't see any use for this, but it makes
174        things uniform.
175    
176    36. Perl no longer allows group names to start with digits, so I have made this
177        change also in PCRE. It simplifies the code a bit.
178    
179    37. In extended mode, Perl ignores spaces before a + that indicates a
180        possessive quantifier. PCRE allowed a space before the quantifier, but not
181        before the possessive +. It now does.
182    
183    38. The use of \K (reset reported match start) within a repeated possessive
184        group such as (a\Kb)*+ was not working.
185    
186    40. Document that the same character tables must be used at compile time and
187        run time, and that the facility to pass tables to pcre_exec() and
188        pcre_dfa_exec() is for use only with saved/restored patterns.
189    
190    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
191        features for Builds with MSVC:
192    
193        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
194           builds). The .rc files can be used to set FileDescription and many other
195           attributes.
196    
197        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
198           This allows higher-level build scripts which want .pdb files to avoid
199           hard-coding the exact files needed."
200    
201    
202    Version 8.33 28-May-2013
203    ------------------------
204    
205    1.  Added 'U' to some constants that are compared to unsigned integers, to
206        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
207        variables that are added to signed variables, to ensure the result is
208        signed and can be negated.
209    
210    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
211        CMake config files.
212    
213    3.  Revise the creation of config.h.generic so that all boolean macros are
214        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
215        overriding via -D on the command line possible.
216    
217    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
218        to unsigned int is reported to make a quite noticeable speed difference in
219        a specific Windows environment. Testing on Linux did also appear to show
220        some benefit (and it is clearly not harmful). Also fixed the definition of
221        Xop which should be unsigned.
222    
223    5.  Related to (4), changing the definition of the intermediate variable cc
224        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
225        improvements.
226    
227    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
228        unnecessary spaces.
229    
230    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
231        and later.
232    
233    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
234    
235    9.  Optimizing fast_forward_start_bits in JIT.
236    
237    10. Adding support for callouts in JIT, and fixing some issues revealed
238        during this work. Namely:
239    
240        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
241    
242        (b) Minimum length was not checked before the matching is started.
243    
244    11. The value of capture_last that is passed to callouts was incorrect in some
245        cases when there was a capture on one path that was subsequently abandoned
246        after a backtrack. Also, the capture_last value is now reset after a
247        recursion, since all captures are also reset in this case.
248    
249    12. The interpreter no longer returns the "too many substrings" error in the
250        case when an overflowing capture is in a branch that is subsequently
251        abandoned after a backtrack.
252    
253    13. In the pathological case when an offset vector of size 2 is used, pcretest
254        now prints out the matched string after a yield of 0 or 1.
255    
256    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
257        Only supported by the JIT compiler at the moment.
258    
259    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
260    
261    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
262        offset of the starting point of the matching process, provided the offsets
263        vector is large enough.
264    
265    17. The \A escape now records a lookbehind value of 1, though its execution
266        does not actually inspect the previous character. This is to ensure that,
267        in partial multi-segment matching, at least one character from the old
268        segment is retained when a new segment is processed. Otherwise, if there
269        are no lookbehinds in the pattern, \A might match incorrectly at the start
270        of a new segment.
271    
272    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
273    
274    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
275        gives some modest performance improvement in 8-bit mode.
276    
277    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
278        be expressed in certain programming languages using Universal Character
279        Names.
280    
281    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
282        which points out that "non characters" are not "characters that may not
283        appear in Unicode strings" but rather "characters that are reserved for
284        internal use and have only local meaning".
285    
286    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
287        there was a conditional group that depended on an assertion, if the
288        assertion was false, the callout that immediately followed the alternation
289        in the condition was skipped when pcre_exec() was used for matching.
290    
291    23. Allow an explicit callout to be inserted before an assertion that is the
292        condition for a conditional group, for compatibility with automatic
293        callouts, which always insert a callout at this point.
294    
295    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
296        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
297    
298    25. (*PRUNE) is now supported by the JIT compiler.
299    
300    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
301    
302    27. Fix the case where there are two or more SKIPs with arguments that may be
303        ignored.
304    
305    28. (*SKIP) is now supported by the JIT compiler.
306    
307    29. (*THEN) is now supported by the JIT compiler.
308    
309    30. Update RunTest with additional test selector options.
310    
311    31. The way PCRE handles backtracking verbs has been changed in two ways.
312    
313        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
314        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
315        backtracking. In some cases this makes it more Perl-compatible, but Perl's
316        rather obscure rules do not always do the same thing.
317    
318        (2) Previously, backtracking verbs were confined within assertions. This is
319        no longer the case for positive assertions, except for (*ACCEPT). Again,
320        this sometimes improves Perl compatibility, and sometimes does not.
321    
322    32. A number of tests that were in test 2 because Perl did things differently
323        have been moved to test 1, because either Perl or PCRE has changed, and
324        these tests are now compatible.
325    
326    32. Backtracking control verbs are now handled in the same way in JIT and
327        interpreter.
328    
329    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
330        contained a forward subroutine reference caused a compile error.
331    
332    34. Auto-detect and optimize limited repetitions in JIT.
333    
334    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
335        blocking (*UTF) etc.
336    
337    36. In the interpreter, maximizing pattern repetitions for characters and
338        character types now use tail recursion, which reduces stack usage.
339    
340    37. The value of the max lookbehind was not correctly preserved if a compiled
341        and saved regex was reloaded on a host of different endianness.
342    
343    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
344        of the compiled pattern block, expand the flags field from 16 to 32 bits
345        because it was almost full.
346    
347    39. Try madvise first before posix_madvise.
348    
349    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
350        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
351        original change locked out all matches of empty strings. This has been
352        changed so that one match of an empty string per line is recognized.
353        Subsequent searches on the same line (for colouring or for --only-matching,
354        for example) do not recognize empty strings.
355    
356    41. Applied a user patch to fix a number of spelling mistakes in comments.
357    
358    42. Data lines longer than 65536 caused pcretest to crash.
359    
360    43. Clarified the data type for length and startoffset arguments for pcre_exec
361        and pcre_dfa_exec in the function-specific man pages, where they were
362        explicitly stated to be in bytes, never having been updated. I also added
363        some clarification to the pcreapi man page.
364    
365    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
366        a segmentation fault.
367    
368    
369    Version 8.32 30-November-2012
370    -----------------------------
371    
372    1.  Improved JIT compiler optimizations for first character search and single
373        character iterators.
374    
375    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
376        Patch by Daniel Richard G.
377    
378    3.  Single character iterator optimizations in the JIT compiler.
379    
380    4.  Improved JIT compiler optimizations for character ranges.
381    
382    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
383        Reported by Giuseppe D'Angelo.
384    
385    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
386        of a line, was being set incorrectly in cases where .* appeared inside
387        atomic brackets at the start of a pattern, or where there was a subsequent
388        *PRUNE or *SKIP.
389    
390    7.  Improved instruction cache flush for POWER/PowerPC.
391        Patch by Daniel Richard G.
392    
393    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
394        grep:
395    
396        (a) There is now no limit to the number of patterns to be matched.
397    
398        (b) An error is given if a pattern is too long.
399    
400        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
401            are now supported.
402    
403        (d) --exclude-from and --include-from (multiple use) have been added.
404    
405        (e) Exclusions and inclusions now apply to all files and directories, not
406            just to those obtained from scanning a directory recursively.
407    
408        (f) Multiple uses of -f and --file-list are now supported.
409    
410        (g) In a Windows environment, the default for -d has been changed from
411            "read" (the GNU grep default) to "skip", because otherwise the presence
412            of a directory in the file list provokes an error.
413    
414        (h) The documentation has been revised and clarified in places.
415    
416    9.  Improve the matching speed of capturing brackets.
417    
418    10. Changed the meaning of \X so that it now matches a Unicode extended
419        grapheme cluster.
420    
421    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
422        out POSIX threads when JIT support is configured.
423    
424    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
425    
426    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
427        structure could go wrong in environments where size_t is not the same size
428        as int.
429    
430    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
431        set.
432    
433    15. The EBCDIC support had decayed; later updates to the code had included
434        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
435        general tidy up of EBCDIC-related issues, and the documentation was also
436        not quite right. There is now a test that can be run on ASCII systems to
437        check some of the EBCDIC-related things (but is it not a full test).
438    
439    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
440        in a small tidy to the code.
441    
442    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
443    
444    18. If the --only-matching (-o) option in pcregrep is specified multiple
445        times, each one causes appropriate output. For example, -o1 -o2 outputs the
446        substrings matched by the 1st and 2nd capturing parentheses. A separating
447        string can be specified by --om-separator (default empty).
448    
449    19. Improving the first n character searches.
450    
451    20. Turn case lists for horizontal and vertical white space into macros so that
452        they are defined only once.
453    
454    21. This set of changes together give more compatible Unicode case-folding
455        behaviour for characters that have more than one other case when UCP
456        support is available.
457    
458        (a) The Unicode property table now has offsets into a new table of sets of
459            three or more characters that are case-equivalent. The MultiStage2.py
460            script that generates these tables (the pcre_ucd.c file) now scans
461            CaseFolding.txt instead of UnicodeData.txt for character case
462            information.
463    
464        (b) The code for adding characters or ranges of characters to a character
465            class has been abstracted into a generalized function that also handles
466            case-independence. In UTF-mode with UCP support, this uses the new data
467            to handle characters with more than one other case.
468    
469        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
470            whose other case is greater than 256 are now correctly matched
471            caselessly. Previously, the high codepoint matched the low one, but not
472            vice versa.
473    
474        (d) The processing of \h, \H, \v, and \ in character classes now makes use
475            of the new class addition function, using character lists defined as
476            macros alongside the case definitions of 20 above.
477    
478        (e) Caseless back references now work with characters that have more than
479            one other case.
480    
481        (f) General caseless matching of characters with more than one other case
482            is supported.
483    
484    22. Unicode character properties were updated from Unicode 6.2.0
485    
486    23. Improved CMake support under Windows. Patch by Daniel Richard G.
487    
488    24. Add support for 32-bit character strings, and UTF-32
489    
490    25. Major JIT compiler update (code refactoring and bugfixing).
491        Experimental Sparc 32 support is added.
492    
493    26. Applied a modified version of Daniel Richard G's patch to create
494        pcre.h.generic and config.h.generic by "make" instead of in the
495        PrepareRelease script.
496    
497    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
498        pcre_compile.c when checking for a zero character.
499    
500    28. Introducing a native interface for JIT. Through this interface, the compiled
501        machine code can be directly executed. The purpose of this interface is to
502        provide fast pattern matching, so several sanity checks are not performed.
503        However, feature tests are still performed. The new interface provides
504        1.4x speedup compared to the old one.
505    
506    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
507        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
508        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
509    
510    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
511        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
512        the "old" RFC 2279). Instead, it ended up passing a negative length to
513        pcre_exec().
514    
515    31. Add support for GCC's visibility feature to hide internal functions.
516    
517    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
518        "unknown -C option" after outputting 0 or 1.
519    
520    33. There is now support for generating a code coverage report for the test
521        suite in environments where gcc is the compiler and lcov is installed. This
522        is mainly for the benefit of the developers.
523    
524    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
525        unaddressable using valgrind annotations, allowing valgrind to detect
526        invalid memory accesses. This is mainly for the benefit of the developers.
527    
528    25. (*UTF) can now be used to start a pattern in any of the three libraries.
529    
530    26. Give configure error if --enable-cpp but no C++ compiler found.
531    
532    
533    Version 8.31 06-July-2012
534    -------------------------
535    
536    1.  Fixing a wrong JIT test case and some compiler warnings.
537    
538    2.  Removed a bashism from the RunTest script.
539    
540    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
541        to unsigned type, result still unsigned" that was given by an MS compiler
542        on encountering the code "-sizeof(xxx)".
543    
544    4.  Partial matching support is added to the JIT compiler.
545    
546    5.  Fixed several bugs concerned with partial matching of items that consist
547        of more than one character:
548    
549        (a) /^(..)\1/ did not partially match "aba" because checking references was
550            done on an "all or nothing" basis. This also applied to repeated
551            references.
552    
553        (b) \R did not give a hard partial match if \r was found at the end of the
554            subject.
555    
556        (c) \X did not give a hard partial match after matching one or more
557            characters at the end of the subject.
558    
559        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
560            a partial match for the string "\r".
561    
562        (e) When newline was set to CRLF, the metacharacter "." did not recognize
563            a partial match for a CR character at the end of the subject string.
564    
565    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
566        running pcretest, the text "(JIT)" added to the output whenever JIT is
567        actually used to run the match.
568    
569    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
570        or /S+[+] with a digit between 1 and 7.
571    
572    8.  OP_NOT now supports any UTF character not just single-byte ones.
573    
574    9.  (*MARK) control verb is now supported by the JIT compiler.
575    
576    10. The command "./RunTest list" lists the available tests without actually
577        running any of them. (Because I keep forgetting what they all are.)
578    
579    11. Add PCRE_INFO_MAXLOOKBEHIND.
580    
581    12. Applied a (slightly modified) user-supplied patch that improves performance
582        when the heap is used for recursion (compiled with --disable-stack-for-
583        recursion). Instead of malloc and free for each heap frame each time a
584        logical recursion happens, frames are retained on a chain and re-used where
585        possible. This sometimes gives as much as 30% improvement.
586    
587    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
588        call.
589    
590    14. As documented, (*COMMIT) is now confined to within a positive assertion.
591    
592    15. It is now possible to link pcretest with libedit as an alternative to
593        libreadline.
594    
595    16. (*COMMIT) control verb is now supported by the JIT compiler.
596    
597    17. The Unicode data tables have been updated to Unicode 6.1.0.
598    
599    18. Added --file-list option to pcregrep.
600    
601    19. Added binary file support to pcregrep, including the -a, --binary-files,
602        -I, and --text options.
603    
604    20. The madvise function is renamed for posix_madvise for QNX compatibility
605        reasons. Fixed by Giuseppe D'Angelo.
606    
607    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
608    
609    22. Changed the option for creating symbolic links for 16-bit man pages from
610        -s to -sf so that re-installing does not cause issues.
611    
612    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
613    
614    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
615        matches in certain environments (the workspace was not being correctly
616        retained). Also added to pcre_dfa_exec() a simple plausibility check on
617        some of the workspace data at the beginning of a restart.
618    
619    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
620        was not doing so when it should - probably a typo introduced by SVN 528
621        (change 8.10/14).
622    
623    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
624        \w+ when the character tables indicated that \x{c4} was a word character.
625        There were several related cases, all because the tests for doing a table
626        lookup were testing for characters less than 127 instead of 255.
627    
628    27. If a pattern contains capturing parentheses that are not used in a match,
629        their slots in the ovector are set to -1. For those that are higher than
630        any matched groups, this happens at the end of processing. In the case when
631        there were back references that the ovector was too small to contain
632        (causing temporary malloc'd memory to be used during matching), and the
633        highest capturing number was not used, memory off the end of the ovector
634        was incorrectly being set to -1. (It was using the size of the temporary
635        memory instead of the true size.)
636    
637    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
638        ovector size, it uses memory at the end of the block that it has got.
639    
640    29. Check for an overlong MARK name and give an error at compile time. The
641        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
642    
643    30. JIT compiler update.
644    
645    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
646        Rill for the patch.
647    
648    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
649    
650    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
651    
652    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
653        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
654        was enabled.
655    
656    35. Improve JIT code generation for greedy plus quantifier.
657    
658    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
659        1 to "aa" instead of to an empty string. The bug affected repeated groups
660        that could potentially match an empty string.
661    
662    37. Optimizing single character iterators in JIT.
663    
664    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
665        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
666        codepoints that are too big for the mode are faulted, and in a UTF mode,
667        disallowed codepoints are also faulted.
668    
669    39. If PCRE was compiled with UTF support, in three places in the DFA
670        matcher there was code that should only have been obeyed in UTF mode, but
671        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
672        processing when bytes with values greater than 127 were present. In 16-bit
673        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
674        both cases the values are those that cannot be the first data item in a UTF
675        character. The three items that might have provoked this were recursions,
676        possessively repeated groups, and atomic groups.
677    
678    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
679        and pcregrep, because some OS require shared objects to be explicitly
680        passed to ld, causing the link step to fail if they are not.
681    
682    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
683        mode, patterns that started with \h* or \R* might be incorrectly matched.
684    
685    
686    Version 8.30 04-February-2012
687    -----------------------------
688    
689    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
690        name is defined in ctype.h.
691    
692    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
693        only in quite long subpatterns.
694    
695    3.  Removed the function pcre_info(), which has been obsolete and deprecated
696        since it was replaced by pcre_fullinfo() in February 2000.
697    
698    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
699        match a (*MARK), and the match failed at the start of the subject, a
700        reference to memory before the start of the subject could occur. This bug
701        was introduced by fix 17 of release 8.21.
702    
703    5.  A reference to an unset group with zero minimum repetition was giving
704        totally wrong answers (in non-JavaScript-compatibility mode). For example,
705        /(another)?(\1?)test/ matched against "hello world test". This bug was
706        introduced in release 8.13.
707    
708    6.  Add support for 16-bit character strings (a large amount of work involving
709        many changes and refactorings).
710    
711    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
712        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
713        from a file.
714    
715    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
716        rounding is not applied in this particular case).
717    
718    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
719        if they appear, or are escaped, in patterns.
720    
721    10. Get rid of a number of -Wunused-but-set-variable warnings.
722    
723    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
724        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
725        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
726        also returns the mark "x". This bug applied to capturing parentheses,
727        non-capturing parentheses, and atomic parentheses. It also applied to some
728        assertions.
729    
730    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
731        information out of configure.ac instead of relying on pcre.h.generic, which
732        is not stored in the repository.
733    
734    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
735        -lreadline.
736    
737    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
738    
739    15. Applied Graycode's patch to put the top-level frame on the stack rather
740        than the heap when not using the stack for recursion. This gives a
741        performance improvement in many cases when recursion is not deep.
742    
743    16. Experimental code added to "pcretest -C" to output the stack frame size.
744    
745    
746    Version 8.21 12-Dec-2011
747    ------------------------
748    
749    1.  Updating the JIT compiler.
750    
751    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
752        are added as well.
753    
754    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
755        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
756        calling _pcre_jit_exec. Some extra comments are added.
757    
758    4.  (*MARK) settings inside atomic groups that do not contain any capturing
759        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
760        was introduced by change 18 for 8.20.
761    
762    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
763        ECMA-262 standard.
764    
765    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
766        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
767        This bug was probably introduced by change 9 of 8.13.
768    
769    7.  While fixing 6 above, I noticed that a number of other items were being
770        incorrectly rejected as "not fixed length". This arose partly because newer
771        opcodes had not been added to the fixed-length checking code. I have (a)
772        corrected the bug and added tests for these items, and (b) arranged for an
773        error to occur if an unknown opcode is encountered while checking for fixed
774        length instead of just assuming "not fixed length". The items that were
775        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
776        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
777        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
778    
779    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
780        being incorrectly compiled and would have given unpredicatble results.
781    
782    9.  A possessively repeated subpattern with minimum repeat count greater than
783        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
784        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
785        the first (A) could occur when it should not.
786    
787    10. Add a cast and remove a redundant test from the code.
788    
789    11. JIT should use pcre_malloc/pcre_free for allocation.
790    
791    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
792        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
793        is anything other than /usr, -L is still shown).
794    
795    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
796    
797    14. Perl does not support \N without a following name in a [] class; PCRE now
798        also gives an error.
799    
800    15. If a forward reference was repeated with an upper limit of around 2000,
801        it caused the error "internal error: overran compiling workspace". The
802        maximum number of forward references (including repeats) was limited by the
803        internal workspace, and dependent on the LINK_SIZE. The code has been
804        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
805        the default depends on LINK_SIZE. There is a new upper limit (for safety)
806        of around 200,000 forward references. While doing this, I also speeded up
807        the filling in of repeated forward references.
808    
809    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
810        incorrectly expecting the subject to contain another "a" after the start.
811    
812    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
813        in the match, the SKIP should be ignored. This was not happening; instead
814        the SKIP was being treated as NOMATCH. For patterns such as
815        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
816        tested.
817    
818    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
819        now much more compatible with Perl, in particular in cases where the result
820        is a non-match for a non-anchored pattern. For example, if
821        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
822        "m", where previously it did not return a name. A side effect of this
823        change is that for partial matches, the last encountered mark name is
824        returned, as for non matches. A number of tests that were previously not
825        Perl-compatible have been moved into the Perl-compatible test files. The
826        refactoring has had the pleasing side effect of removing one argument from
827        the match() function, thus reducing its stack requirements.
828    
829    19. If the /S+ option was used in pcretest to study a pattern using JIT,
830        subsequent uses of /S (without +) incorrectly behaved like /S+.
831    
832    21. Retrieve executable code size support for the JIT compiler and fixing
833        some warnings.
834    
835    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
836        not work when the shorter character appeared right at the end of the
837        subject string.
838    
839    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
840        systems.
841    
842    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
843        output it when the /M option is used in pcretest.
844    
845    25. The CheckMan script was not being included in the distribution. Also, added
846        an explicit "perl" to run Perl scripts from the PrepareRelease script
847        because this is reportedly needed in Windows.
848    
849    26. If study data was being save in a file and studying had not found a set of
850        "starts with" bytes for the pattern, the data written to the file (though
851        never used) was taken from uninitialized memory and so caused valgrind to
852        complain.
853    
854    27. Updated RunTest.bat as provided by Sheri Pierce.
855    
856    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
857    
858    29. Computation of memory usage for the table of capturing group names was
859        giving an unnecessarily large value.
860    
861    
862    Version 8.20 21-Oct-2011
863    ------------------------
864    
865    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
866        a POSIX class. After further experiments with Perl, which convinced me that
867        Perl has bugs and confusions, a closing square bracket is no longer allowed
868        in a POSIX name. This bug also affected patterns with classes that started
869        with full stops.
870    
871    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
872        captured substring, but while checking the failing first alternative,
873        substring 1 is temporarily captured. If the output vector supplied to
874        pcre_exec() was not big enough for this capture, the yield of the function
875        was still zero ("insufficient space for captured substrings"). This cannot
876        be totally fixed without adding another stack variable, which seems a lot
877        of expense for a edge case. However, I have improved the situation in cases
878        such as /(a)(b)x|abc/ matched against "abc", where the return code
879        indicates that fewer than the maximum number of slots in the ovector have
880        been set.
881    
882    3.  Related to (2) above: when there are more back references in a pattern than
883        slots in the output vector, pcre_exec() uses temporary memory during
884        matching, and copies in the captures as far as possible afterwards. It was
885        using the entire output vector, but this conflicts with the specification
886        that only 2/3 is used for passing back captured substrings. Now it uses
887        only the first 2/3, for compatibility. This is, of course, another edge
888        case.
889    
890    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
891        main code base, and can be used by building with --enable-jit. When this is
892        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
893        runtime --no-jit option is given.
894    
895    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
896        ovector, the return from the function was zero, implying that there were
897        other matches that did not fit. The correct "exactly full" value is now
898        returned.
899    
900    6.  If a subpattern that was called recursively or as a subroutine contained
901        (*PRUNE) or any other control that caused it to give a non-standard return,
902        invalid errors such as "Error -26 (nested recursion at the same subject
903        position)" or even infinite loops could occur.
904    
905    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
906        computing the minimum length on reaching *ACCEPT, and so ended up with the
907        wrong value of 1 rather than 0. Further investigation indicates that
908        computing a minimum subject length in the presence of *ACCEPT is difficult
909        (think back references, subroutine calls), and so I have changed the code
910        so that no minimum is registered for a pattern that contains *ACCEPT.
911    
912    8.  If (*THEN) was present in the first (true) branch of a conditional group,
913        it was not handled as intended. [But see 16 below.]
914    
915    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
916        Sheri Pierce.
917    
918    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
919        the first byte in a match must be "a".
920    
921    11. Change 17 for 8.13 increased the recursion depth for patterns like
922        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
923        pattern contains any instances of (*THEN). If it does not, the old
924        optimizations are restored. It would be nice to do this on a per-group
925        basis, but at the moment that is not feasible.
926    
927    12. In some environments, the output of pcretest -C is CRLF terminated. This
928        broke RunTest's code that checks for the link size. A single white space
929        character after the value is now allowed for.
930    
931    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
932        For "fr", it uses the Windows-specific input and output files.
933    
934    14. If (*THEN) appeared in a group that was called recursively or as a
935        subroutine, it did not work as intended. [But see next item.]
936    
937    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
938        pattern fragments (but not containing any | characters). If A and B are
939        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
940        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
941        D. In other words, Perl considers parentheses that do not contain any |
942        characters to be part of a surrounding alternative, whereas PCRE was
943        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
944        differently. PCRE now behaves in the same way as Perl, except in the case
945        of subroutine/recursion calls such as (?1) which have in any case always
946        been different (but PCRE had them first :-).
947    
948    16. Related to 15 above: Perl does not treat the | in a conditional group as
949        creating alternatives. Such a group is treated in the same way as an
950        ordinary group without any | characters when processing (*THEN). PCRE has
951        been changed to match Perl's behaviour.
952    
953    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
954        RunGrepTest script failed.
955    
956    18. Change 22 for version 13 caused atomic groups to use more stack. This is
957        inevitable for groups that contain captures, but it can lead to a lot of
958        stack use in large patterns. The old behaviour has been restored for atomic
959        groups that do not contain any capturing parentheses.
960    
961    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
962        suppress the check for a minimum subject length at run time. (If it was
963        given to pcre_exec() or pcre_dfa_exec() it did work.)
964    
965    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
966        fail to work when decoding hex characters in data strings in EBCDIC
967        environments.
968    
969    21. It appears that in at least one Mac OS environment, the isxdigit() function
970        is implemented as a macro that evaluates to its argument more than once,
971        contravening the C 90 Standard (I haven't checked a later standard). There
972        was an instance in pcretest which caused it to go wrong when processing
973        \x{...} escapes in subject strings. The has been rewritten to avoid using
974        things like p++ in the argument of isxdigit().
975    
976    
977    Version 8.13 16-Aug-2011
978    ------------------------
979    
980    1.  The Unicode data tables have been updated to Unicode 6.0.0.
981    
982    2.  Two minor typos in pcre_internal.h have been fixed.
983    
984    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
985        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
986        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
987    
988    4.  There were a number of related bugs in the code for matching backrefences
989        caselessly in UTF-8 mode when codes for the characters concerned were
990        different numbers of bytes. For example, U+023A and U+2C65 are an upper
991        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
992        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
993        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
994        2-byte code at the end of the subject (it thought there wasn't enough data
995        left).
996    
997    5.  Comprehensive information about what went wrong is now returned by
998        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
999        as the output vector has at least 2 elements. The offset of the start of
1000        the failing character and a reason code are placed in the vector.
1001    
1002    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1003        now returned is for the first byte of the failing character, instead of the
1004        last byte inspected. This is an incompatible change, but I hope it is small
1005        enough not to be a problem. It makes the returned offset consistent with
1006        pcre_exec() and pcre_dfa_exec().
1007    
1008    7.  pcretest now gives a text phrase as well as the error number when
1009        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1010        failure, the offset and reason code are output.
1011    
1012    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1013        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1014        back over a single character (\n). This seems wrong (because it treated the
1015        two characters as a single entity when going forwards), conflicts with the
1016        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1017        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1018        has been changed.
1019    
1020    9.  Some internal refactoring has changed the processing so that the handling
1021        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1022        time (the PCRE_DOTALL option was changed this way some time ago: version
1023        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1024        which was always a bit of a fudge. It also means that there is one less
1025        argument for the match() function, which reduces its stack requirements
1026        slightly. This change also fixes an incompatibility with Perl: the pattern
1027        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1028    
1029    10. More internal refactoring has drastically reduced the number of recursive
1030        calls to match() for possessively repeated groups such as (abc)++ when
1031        using pcre_exec().
1032    
1033    11. While implementing 10, a number of bugs in the handling of groups were
1034        discovered and fixed:
1035    
1036        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1037        (a|)*(?1) gave a compile-time internal error.
1038        ((a|)+)+  did not notice that the outer group could match an empty string.
1039        (^a|^)+   was not marked as anchored.
1040        (.*a|.*)+ was not marked as matching at start or after a newline.
1041    
1042    12. Yet more internal refactoring has removed another argument from the match()
1043        function. Special calls to this function are now indicated by setting a
1044        value in a variable in the "match data" data block.
1045    
1046    13. Be more explicit in pcre_study() instead of relying on "default" for
1047        opcodes that mean there is no starting character; this means that when new
1048        ones are added and accidentally left out of pcre_study(), testing should
1049        pick them up.
1050    
1051    14. The -s option of pcretest has been documented for ages as being an old
1052        synonym of -m (show memory usage). I have changed it to mean "force study
1053        for every regex", that is, assume /S for every regex. This is similar to -i
1054        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1055        using it. It makes it easier to run collections of tests with and without
1056        study enabled, and thereby test pcre_study() more easily. All the standard
1057        tests are now run with and without -s (but some patterns can be marked as
1058        "never study" - see 20 below).
1059    
1060    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1061        restoration of the capturing data to the outer values was not happening
1062        correctly.
1063    
1064    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1065        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1066        pattern had matched an empty string, and so incorrectly returned a no
1067        match.
1068    
1069    17. There was optimizing code for the last branch of non-capturing parentheses,
1070        and also for the obeyed branch of a conditional subexpression, which used
1071        tail recursion to cut down on stack usage. Unfortunately, now that there is
1072        the possibility of (*THEN) occurring in these branches, tail recursion is
1073        no longer possible because the return has to be checked for (*THEN). These
1074        two optimizations have therefore been removed. [But see 8.20/11 above.]
1075    
1076    18. If a pattern containing \R was studied, it was assumed that \R always
1077        matched two bytes, thus causing the minimum subject length to be
1078        incorrectly computed because \R can also match just one byte.
1079    
1080    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1081        was incorrectly computed.
1082    
1083    20. If /S is present twice on a test pattern in pcretest input, it now
1084        *disables* studying, thereby overriding the use of -s on the command line
1085        (see 14 above). This is necessary for one or two tests to keep the output
1086        identical in both cases.
1087    
1088    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1089        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1090    
1091    22. When an atomic group that contained a capturing parenthesis was
1092        successfully matched, but the branch in which it appeared failed, the
1093        capturing was not being forgotten if a higher numbered group was later
1094        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1095        group 1 to "a", when in fact it should be unset. This applied to multi-
1096        branched capturing and non-capturing groups, repeated or not, and also to
1097        positive assertions (capturing in negative assertions does not happen
1098        in PCRE) and also to nested atomic groups.
1099    
1100    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1101        subject after a captured substring, to make it easier to tell which of a
1102        number of identical substrings has been captured.
1103    
1104    24. The way atomic groups are processed by pcre_exec() has been changed so that
1105        if they are repeated, backtracking one repetition now resets captured
1106        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1107        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1108        "aaa". Previously, it would have been "a". As part of this code
1109        refactoring, the way recursive calls are handled has also been changed.
1110    
1111    25. If an assertion condition captured any substrings, they were not passed
1112        back unless some other capturing happened later. For example, if
1113        (?(?=(a))a) was matched against "a", no capturing was returned.
1114    
1115    26. When studying a pattern that contained subroutine calls or assertions,
1116        the code for finding the minimum length of a possible match was handling
1117        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1118        group 1 called group 2 while simultaneously a separate group 2 called group
1119        1). A stack overflow occurred in this case. I have fixed this by limiting
1120        the recursion depth to 10.
1121    
1122    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1123        Fortmann. This supports explicit test numbers on the command line, and has
1124        argument validation and error reporting.
1125    
1126    28. An instance of \X with an unlimited repeat could fail if at any point the
1127        first character it looked at was a mark character.
1128    
1129    29. Some minor code refactoring concerning Unicode properties and scripts
1130        should reduce the stack requirement of match() slightly.
1131    
1132    30. Added the '=' option to pcretest to check the setting of unused capturing
1133        slots at the end of the pattern, which are documented as being -1, but are
1134        not included in the return count.
1135    
1136    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1137        compiled something random. Now it gives a compile-time error (as does
1138        Perl).
1139    
1140    32. A *MARK encountered during the processing of a positive assertion is now
1141        recorded and passed back (compatible with Perl).
1142    
1143    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1144        had alternative anchored branches, the search for a second match in a line
1145        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1146        matched the line "0102" twice. The same bug affected patterns that started
1147        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1148        twice.
1149    
1150    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1151        does, and because of capturing effects, quantifying parenthesized
1152        assertions may at times be useful. Quantifiers are now allowed for
1153        parenthesized assertions.
1154    
1155    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1156    
1157    36. \g was being checked for fancy things in a character class, when it should
1158        just be a literal "g".
1159    
1160    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1161        appearance of a nested POSIX class supersedes an apparent external class.
1162        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1163        unescaped square brackets may also appear as part of class names. For
1164        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1165        more like Perl. (But see 8.20/1 above.)
1166    
1167    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1168        was because it thought it was \N{name}, which is not supported).
1169    
1170    39. Add minix to OS list not supporting the -S option in pcretest.
1171    
1172    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1173        cannot analyze patterns in sufficient detail to catch mutual recursions
1174        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1175        subgroup is called recursively as a subpattern for a second time at the
1176        same position in the subject string. In previous releases this might have
1177        been caught by the recursion limit, or it might have run out of stack.
1178    
1179    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1180        happen only once. PCRE was, however incorrectly giving a compile time error
1181        "recursive call could loop indefinitely" because it cannot analyze the
1182        pattern in sufficient detail. The compile time test no longer happens when
1183        PCRE is compiling a conditional subpattern, but actual runaway loops are
1184        now caught at runtime (see 40 above).
1185    
1186    42. It seems that Perl allows any characters other than a closing parenthesis
1187        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1188        has been changed to be the same.
1189    
1190    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1191        as not to get warnings when autogen.sh is called. Also changed
1192        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1193    
1194    44. To help people who use pcregrep to scan files containing exceedingly long
1195        lines, the following changes have been made:
1196    
1197        (a) The default value of the buffer size parameter has been increased from
1198            8K to 20K. (The actual buffer used is three times this size.)
1199    
1200        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1201            PCRE is built.
1202    
1203        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1204            to be set at run time.
1205    
1206        (d) Numerical values in pcregrep options can be followed by K or M, for
1207            example --buffer-size=50K.
1208    
1209        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1210            given and the return code is set to 2.
1211    
1212    45. Add a pointer to the latest mark to the callout data block.
1213    
1214    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1215        partial match of an empty string instead of no match. This was specific to
1216        the use of ".".
1217    
1218    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1219        complete match instead of a partial match. This bug was dependent on both
1220        the PCRE_UTF8 and PCRE_DOTALL options being set.
1221    
1222    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1223        starting byte set, because \b was not being ignored.
1224    
1225    
1226    Version 8.12 15-Jan-2011
1227    ------------------------
1228    
1229    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1230        checks for such things as part of the documentation building process.
1231    
1232    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1233        --match-limit and --recursion-limit options (added for 8.11). In
1234        particular, this made one of the standard tests fail. (The integer value
1235        went into the wrong half of a long int.)
1236    
1237    3.  If the --colour option was given to pcregrep with -v (invert match), it
1238        did strange things, either producing crazy output, or crashing. It should,
1239        of course, ignore a request for colour when reporting lines that do not
1240        match.
1241    
1242    4.  Another pcregrep bug caused similar problems if --colour was specified with
1243        -M (multiline) and the pattern match finished with a line ending.
1244    
1245    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1246        matched in multiline mode, the following line was shown as part of the
1247        match. This seems wrong, so I have changed it.
1248    
1249    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1250        the check for further matches in the same line (so they could be coloured)
1251        to overrun the end of the current line. If another match was found, it was
1252        incorrectly shown (and then shown again when found in the next line).
1253    
1254    7.  If pcregrep was compiled under Windows, there was a reference to the
1255        function pcregrep_exit() before it was defined. I am assuming this was
1256        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1257        reported by a user. I've moved the definition above the reference.
1258    
1259    
1260    Version 8.11 10-Dec-2010
1261    ------------------------
1262    
1263    1.  (*THEN) was not working properly if there were untried alternatives prior
1264        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1265        backtracked to try for "b" instead of moving to the next alternative branch
1266        at the same level (in this case, to look for "c"). The Perl documentation
1267        is clear that when (*THEN) is backtracked onto, it goes to the "next
1268        alternative in the innermost enclosing group".
1269    
1270    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1271        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1272        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1273        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1274        (*THEN).
1275    
1276    3.  If \s appeared in a character class, it removed the VT character from
1277        the class, even if it had been included by some previous item, for example
1278        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1279        of \s, but is part of the POSIX "space" class.)
1280    
1281    4.  A partial match never returns an empty string (because you can always
1282        match an empty string at the end of the subject); however the checking for
1283        an empty string was starting at the "start of match" point. This has been
1284        changed to the "earliest inspected character" point, because the returned
1285        data for a partial match starts at this character. This means that, for
1286        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1287        (previously it gave "no match").
1288    
1289    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1290        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1291        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1292        has an implication that the given string is incomplete (because a partial
1293        match is preferred over a full match). For this reason, these items now
1294        give a partial match in this situation. [Aside: previously, the one case
1295        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1296        match rather than a full match, which was wrong by the old rules, but is
1297        now correct.]
1298    
1299    6.  There was a bug in the handling of #-introduced comments, recognized when
1300        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1301        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1302        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1303        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1304        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1305        places in pcre_compile().
1306    
1307    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1308        comments when looking ahead for named forward references to subpatterns,
1309        the only newline sequence it recognized was NL. It now handles newlines
1310        according to the set newline convention.
1311    
1312    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1313        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1314        cater for a lack of strerror(). These oversights have been fixed.
1315    
1316    9.  Added --match-limit and --recursion-limit to pcregrep.
1317    
1318    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1319    
1320    11. When the -o option was used, pcregrep was setting a return code of 1, even
1321        when matches were found, and --line-buffered was not being honoured.
1322    
1323    12. Added an optional parentheses number to the -o and --only-matching options
1324        of pcregrep.
1325    
1326    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1327        can match an empty string. The code to do it in pcretest and pcredemo
1328        needed fixing:
1329    
1330        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1331            only one byte after an empty string match just before CRLF (this case
1332            just got forgotten; "any" and "anycrlf" were OK).
1333    
1334        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1335            mode when an empty string match preceded an ASCII character followed by
1336            a non-ASCII character. (The code for advancing by one character rather
1337            than one byte was nonsense.)
1338    
1339        (c) The pcredemo.c sample program did not have any code at all to handle
1340            the cases when CRLF is a valid newline sequence.
1341    
1342    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1343        as a starting offset was within the subject string. There is now a new
1344        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1345        negative or greater than the length of the string. In order to test this,
1346        pcretest is extended to allow the setting of negative starting offsets.
1347    
1348    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1349        starting offset points to the beginning of a UTF-8 character was
1350        unnecessarily clumsy. I tidied it up.
1351    
1352    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1353        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1354    
1355    17. Nobody had reported that the --include_dir option, which was added in
1356        release 7.7 should have been called --include-dir (hyphen, not underscore)
1357        for compatibility with GNU grep. I have changed it to --include-dir, but
1358        left --include_dir as an undocumented synonym, and the same for
1359        --exclude-dir, though that is not available in GNU grep, at least as of
1360        release 2.5.4.
1361    
1362    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1363        characters from a string of bytes) have been redefined so as not to use
1364        loops, in order to improve performance in some environments. At the same
1365        time, I abstracted some of the common code into auxiliary macros to save
1366        repetition (this should not affect the compiled code).
1367    
1368    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1369        compile-time error is now given if \c is not followed by an ASCII
1370        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1371        different, and any byte value is allowed.)
1372    
1373    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1374        START_OPTIMIZE option, which is now allowed at compile time - but just
1375        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1376        to pcregrep and other applications that have no direct access to PCRE
1377        options. The new /Y option in pcretest sets this option when calling
1378        pcre_compile().
1379    
1380    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1381        back references. Groups containing recursive back references were forced to
1382        be atomic by that change, but in the case of named groups, the amount of
1383        memory required was incorrectly computed, leading to "Failed: internal
1384        error: code overflow". This has been fixed.
1385    
1386    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1387        pcretest.c, to avoid build problems in some Borland environments.
1388    
1389    
1390    Version 8.10 25-Jun-2010
1391    ------------------------
1392    
1393    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1394        THEN.
1395    
1396    2.  (*ACCEPT) was not working when inside an atomic group.
1397    
1398    3.  Inside a character class, \B is treated as a literal by default, but
1399        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1400        causes the error). The code is unchanged, but I tidied the documentation.
1401    
1402    4.  Inside a character class, PCRE always treated \R and \X as literals,
1403        whereas Perl faults them if its -w option is set. I have changed PCRE so
1404        that it faults them when PCRE_EXTRA is set.
1405    
1406    5.  Added support for \N, which always matches any character other than
1407        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1408    
1409    6.  When compiling pcregrep with newer versions of gcc which may have
1410        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1411        declared with attribute warn_unused_result" were given. Just casting the
1412        result to (void) does not stop the warnings; a more elaborate fudge is
1413        needed. I've used a macro to implement this.
1414    
1415    7.  Minor change to pcretest.c to avoid a compiler warning.
1416    
1417    8.  Added four artifical Unicode properties to help with an option to make
1418        \s etc use properties (see next item). The new properties are: Xan
1419        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1420    
1421    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1422        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1423        this option. Modified pcretest to add /W to test this facility. Added
1424        REG_UCP to make it available via the POSIX interface.
1425    
1426    10. Added --line-buffered to pcregrep.
1427    
1428    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1429        studied, and the match started with a letter with a code point greater than
1430        127 whose first byte was different to the first byte of the other case of
1431        the letter, the other case of this starting letter was not recognized
1432        (#976).
1433    
1434    12. If a pattern that was studied started with a repeated Unicode property
1435        test, for example, \p{Nd}+, there was the theoretical possibility of
1436        setting up an incorrect bitmap of starting bytes, but fortunately it could
1437        not have actually happened in practice until change 8 above was made (it
1438        added property types that matched character-matching opcodes).
1439    
1440    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1441        possible starting bytes for non-anchored patterns.
1442    
1443    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1444        \R, and also a number of cases that involve Unicode properties, both
1445        explicit and implicit when PCRE_UCP is set.
1446    
1447    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1448        input, it could crash or give wrong results if characters with values
1449        greater than 0xc0 were present in the subject string. (Detail: it assumed
1450        UTF-8 input when processing these items.)
1451    
1452    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1453        size_t is 64-bit (#991).
1454    
1455    17. Added a check for running out of memory when PCRE is compiled with
1456        --disable-stack-for-recursion (#990).
1457    
1458    18. If the last data line in a file for pcretest does not have a newline on
1459        the end, a newline was missing in the output.
1460    
1461    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1462        less than 128) in its various bitmaps. However, there is a facility for
1463        generating tables according to the current locale when PCRE is compiled. It
1464        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1465        characters, are recognized by isspace() and therefore were getting set in
1466        these tables, and indeed these tables seem to approximate to ISO 8859. This
1467        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1468        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1469        which of course cannot start UTF-8 characters. I have changed the code so
1470        that only real ASCII characters (less than 128) and the correct starting
1471        bytes for UTF-8 encodings are set for characters greater than 127 when in
1472        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1473        altogether.)
1474    
1475    20. Added the /T option to pcretest so as to be able to run tests with non-
1476        standard character tables, thus making it possible to include the tests
1477        used for 19 above in the standard set of tests.
1478    
1479    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1480        reference to a subpattern the other side of a comment that contains an
1481        opening parenthesis caused either an internal compiling error, or a
1482        reference to the wrong subpattern.
1483    
1484    
1485    Version 8.02 19-Mar-2010
1486    ------------------------
1487    
1488    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1489    
1490    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1491        configured.
1492    
1493    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1494        original author of that file, following a query about its status.
1495    
1496    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1497        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1498    
1499    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1500        quantifier applied to a forward-referencing subroutine call, could compile
1501        incorrect code or give the error "internal error: previously-checked
1502        referenced subpattern not found".
1503    
1504    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1505        variables to point to external functions. For these systems, therefore,
1506        pcre_malloc etc. are now initialized to local functions that call the
1507        relevant global functions.
1508    
1509    7.  There were two entries missing in the vectors called coptable and poptable
1510        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1511        I've fixed the data, and added a kludgy way of testing at compile time that
1512        the lengths are correct (equal to the number of opcodes).
1513    
1514    8.  Following on from 7, I added a similar kludge to check the length of the
1515        eint vector in pcreposix.c.
1516    
1517    9.  Error texts for pcre_compile() are held as one long string to avoid too
1518        much relocation at load time. To find a text, the string is searched,
1519        counting zeros. There was no check for running off the end of the string,
1520        which could happen if a new error number was added without updating the
1521        string.
1522    
1523    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1524    
1525    11. \K was not working if it appeared in an atomic group or in a group that
1526        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1527        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1528        the assertion is positive, but not if it is negative.
1529    
1530    12. Change 11 fortuitously reduced the size of the stack frame used in the
1531        "match()" function of pcre_exec.c by one pointer. Forthcoming
1532        implementation of support for (*MARK) will need an extra pointer on the
1533        stack; I have reserved it now, so that the stack frame size does not
1534        decrease.
1535    
1536    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1537        item in branch that calls a recursion is a subroutine call - as in the
1538        second branch in the above example - was incorrectly given the compile-
1539        time error "recursive call could loop indefinitely" because pcre_compile()
1540        was not correctly checking the subroutine for matching a non-empty string.
1541    
1542    14. The checks for overrunning compiling workspace could trigger after an
1543        overrun had occurred. This is a "should never occur" error, but it can be
1544        triggered by pathological patterns such as hundreds of nested parentheses.
1545        The checks now trigger 100 bytes before the end of the workspace.
1546    
1547    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1548    
1549    
1550    Version 8.01 19-Jan-2010
1551    ------------------------
1552    
1553    1.  If a pattern contained a conditional subpattern with only one branch (in
1554        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1555        computed the wrong minimum data length (which is of course zero for such
1556        subpatterns). This could cause incorrect "no match" results.
1557    
1558    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1559        the pattern is reset in the first branch, pcre_compile() failed with
1560        "internal error: code overflow at offset...". This happened only when
1561        the reset was to the original external option setting. (An optimization
1562        abstracts leading options settings into an external setting, which was the
1563        cause of this.)
1564    
1565    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1566        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1567        assertion pattern did not match (meaning that the assertion was true), it
1568        was incorrectly treated as false if the SKIP had been reached during the
1569        matching. This also applied to assertions used as conditions.
1570    
1571    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1572        assertion subpattern, including such a pattern used as a condition,
1573        unpredictable results occurred, instead of the error return
1574        PCRE_ERROR_DFA_UITEM.
1575    
1576    5.  The C++ GlobalReplace function was not working like Perl for the special
1577        situation when an empty string is matched. It now does the fancy magic
1578        stuff that is necessary.
1579    
1580    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1581        removed. (These were left over from very, very early versions of PCRE.)
1582    
1583    7.  Some cosmetic changes to the code to make life easier when compiling it
1584        as part of something else:
1585    
1586        (a) Change DEBUG to PCRE_DEBUG.
1587    
1588        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1589            called "current" as "current_branch", to prevent a collision with the
1590            Linux macro when compiled as a kernel module.
1591    
1592        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1593            prevent a collision with the Linux macro when compiled as a kernel
1594            module.
1595    
1596    8.  In pcre_compile() there are some checks for integer overflows that used to
1597        cast potentially large values to (double). This has been changed to that
1598        when building, a check for int64_t is made, and if it is found, it is used
1599        instead, thus avoiding the use of floating point arithmetic. (There is no
1600        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1601        double.
1602    
1603    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1604        2005 (difference between two addresses compared to an unsigned value).
1605    
1606    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1607        custom one, because of the following reported problem in Windows:
1608    
1609          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1610              under Win32.
1611          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1612              therefore missing the function definition.
1613          - The compiler thus generates a "C" signature for the test function.
1614          - The linker fails to find the "C" function.
1615          - PCRE fails to configure if asked to do so against libbz2.
1616    
1617    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1618        messages were output:
1619    
1620          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1621          rerunning libtoolize, to keep the correct libtool macros in-tree.
1622          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1623    
1624        I have done both of these things.
1625    
1626    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1627        most of the time, it *can* run out if it is given a pattern that contains a
1628        runaway infinite recursion. I updated the discussion in the pcrestack man
1629        page.
1630    
1631    13. Now that we have gone to the x.xx style of version numbers, the minor
1632        version may start with zero. Using 08 or 09 is a bad idea because users
1633        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1634        interpreted as invalid octal numbers. I've updated the previous comment in
1635        configure.ac, and also added a check that gives an error if 08 or 09 are
1636        used.
1637    
1638    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1639        causing partial matching to fail when the end of the subject matched \W
1640        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1641    
1642    15. There were some discrepancies between the declarations in pcre_internal.h
1643        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1644        their definitions. The declarations used "const uschar *" and the
1645        definitions used USPTR. Even though USPTR is normally defined as "const
1646        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1647        reported that: "This difference in casting confuses some C++ compilers, for
1648        example, SunCC recognizes above declarations as different functions and
1649        generates broken code for hbpcre." I have changed the declarations to use
1650        USPTR.
1651    
1652    16. GNU libtool is named differently on some systems. The autogen.sh script now
1653        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1654        (FreeBSD).
1655    
1656    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1657        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1658        comment: "Figure out how to create a longlong from a string: strtoll and
1659        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1660        instance, but it only takes 2 args instead of 3!"
1661    
1662    18. A subtle bug concerned with back references has been fixed by a change of
1663        specification, with a corresponding code fix. A pattern such as
1664        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1665        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1666        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1667        same bug. Such groups have to be quantified to be useful, or contained
1668        inside another quantified group. (If there's no repetition, the reference
1669        can never match.) The problem arises because, having left the group and
1670        moved on to the rest of the pattern, a later failure that backtracks into
1671        the group uses the captured value from the final iteration of the group
1672        rather than the correct earlier one. I have fixed this in PCRE by forcing
1673        any group that contains a reference to itself to be an atomic group; that
1674        is, there cannot be any backtracking into it once it has completed. This is
1675        similar to recursive and subroutine calls.
1676    
1677    
1678    Version 8.00 19-Oct-09
1679    ----------------------
1680    
1681    1.  The table for translating pcre_compile() error codes into POSIX error codes
1682        was out-of-date, and there was no check on the pcre_compile() error code
1683        being within the table. This could lead to an OK return being given in
1684        error.
1685    
1686    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1687        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1688        in a Windows environment.
1689    
1690    3.  The pcregrep --count option prints the count for each file even when it is
1691        zero, as does GNU grep. However, pcregrep was also printing all files when
1692        --files-with-matches was added. Now, when both options are given, it prints
1693        counts only for those files that have at least one match. (GNU grep just
1694        prints the file name in this circumstance, but including the count seems
1695        more useful - otherwise, why use --count?) Also ensured that the
1696        combination -clh just lists non-zero counts, with no names.
1697    
1698    4.  The long form of the pcregrep -F option was incorrectly implemented as
1699        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1700        but it seems right to fix it, and I didn't think it was worth preserving
1701        the old behaviour.
1702    
1703    5.  The command line items --regex=pattern and --regexp=pattern were not
1704        recognized by pcregrep, which required --regex pattern or --regexp pattern
1705        (with a space rather than an '='). The man page documented the '=' forms,
1706        which are compatible with GNU grep; these now work.
1707    
1708    6.  No libpcreposix.pc file was created for pkg-config; there was just
1709        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1710    
1711    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1712        when UCP support is not needed, by modifying the Python script that
1713        generates it from Unicode data files. This should not matter if the module
1714        is correctly used as a library, but I received one complaint about 50K of
1715        unwanted data. My guess is that the person linked everything into his
1716        program rather than using a library. Anyway, it does no harm.
1717    
1718    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1719        was a minimum greater than 1 for a wide character in a possessive
1720        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1721        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1722        character. Chaos in the form of incorrect output or a compiling loop could
1723        result.
1724    
1725    9.  The restrictions on what a pattern can contain when partial matching is
1726        requested for pcre_exec() have been removed. All patterns can now be
1727        partially matched by this function. In addition, if there are at least two
1728        slots in the offset vector, the offset of the earliest inspected character
1729        for the match and the offset of the end of the subject are set in them when
1730        PCRE_ERROR_PARTIAL is returned.
1731    
1732    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1733        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1734        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1735        and may be more useful for multi-segment matching.
1736    
1737    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1738        used to be given if ever the end of the subject was reached; now it is
1739        given only if matching could not proceed because another character was
1740        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1741        string "Z", which now yields "no match" instead of "partial match". In the
1742        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1743        final character ended with (*FAIL).
1744    
1745    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1746        if the pattern had a "must contain" character that was already found in the
1747        earlier partial match, unless partial matching was again requested. For
1748        example, with the pattern /dog.(body)?/, the "must contain" character is
1749        "g". If the first part-match was for the string "dog", restarting with
1750        "sbody" failed. This bug has been fixed.
1751    
1752    13. The string returned by pcre_dfa_exec() after a partial match has been
1753        changed so that it starts at the first inspected character rather than the
1754        first character of the match. This makes a difference only if the pattern
1755        starts with a lookbehind assertion or \b or \B (\K is not supported by
1756        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1757        matching functions compatible, and I think it's the right thing to do.
1758    
1759    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1760        so that the demonstration program is easily available in environments where
1761        PCRE has not been installed from source.
1762    
1763    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1764        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1765        library.
1766    
1767    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1768        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1769        is not the first non-POSIX option to be added. Clearly some people find
1770        these options useful.
1771    
1772    17. If a caller to the POSIX matching function regexec() passes a non-zero
1773        value for nmatch with a NULL value for pmatch, the value of
1774        nmatch is forced to zero.
1775    
1776    18. RunGrepTest did not have a test for the availability of the -u option of
1777        the diff command, as RunTest does. It now checks in the same way as
1778        RunTest, and also checks for the -b option.
1779    
1780    19. If an odd number of negated classes containing just a single character
1781        interposed, within parentheses, between a forward reference to a named
1782        subpattern and the definition of the subpattern, compilation crashed with
1783        an internal error, complaining that it could not find the referenced
1784        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1785        [The bug was that it was starting one character too far in when skipping
1786        over the character class, thus treating the ] as data rather than
1787        terminating the class. This meant it could skip too much.]
1788    
1789    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1790        /g option in pcretest when the pattern contains \K, which makes it possible
1791        to have an empty string match not at the start, even when the pattern is
1792        anchored. Updated pcretest and pcredemo to use this option.
1793    
1794    21. If the maximum number of capturing subpatterns in a recursion was greater
1795        than the maximum at the outer level, the higher number was returned, but
1796        with unset values at the outer level. The correct (outer level) value is
1797        now given.
1798    
1799    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1800        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1801        make it do so. The string so far is captured, making this feature
1802        compatible with Perl.
1803    
1804    23. The tests have been re-organized, adding tests 11 and 12, to make it
1805        possible to check the Perl 5.10 features against Perl 5.10.
1806    
1807    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1808        pattern matches a fixed length string. PCRE did not allow this; now it
1809        does. Neither allows recursion.
1810    
1811    25. I finally figured out how to implement a request to provide the minimum
1812        length of subject string that was needed in order to match a given pattern.
1813        (It was back references and recursion that I had previously got hung up
1814        on.) This code has now been added to pcre_study(); it finds a lower bound
1815        to the length of subject needed. It is not necessarily the greatest lower
1816        bound, but using it to avoid searching strings that are too short does give
1817        some useful speed-ups. The value is available to calling programs via
1818        pcre_fullinfo().
1819    
1820    26. While implementing 25, I discovered to my embarrassment that pcretest had
1821        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1822        study optimizations had never been tested with that matching function.
1823        Oops. What is worse, even when it was passed study data, there was a bug in
1824        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1825        were also very few tests of studied patterns with pcre_dfa_exec().
1826    
1827    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1828        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1829        on the other side of the coin, they are no longer allowed to have different
1830        names, because these cannot be distinguished in PCRE, and this has caused
1831        confusion. (This is a difference from Perl.)
1832    
1833    28. When duplicate subpattern names are present (necessarily with different
1834        numbers, as required by 27 above), and a test is made by name in a
1835        conditional pattern, either for a subpattern having been matched, or for
1836        recursion in such a pattern, all the associated numbered subpatterns are
1837        tested, and the overall condition is true if the condition is true for any
1838        one of them. This is the way Perl works, and is also more like the way
1839        testing by number works.
1840    
1841    
1842    Version 7.9 11-Apr-09
1843    ---------------------
1844    
1845    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1846        (pcretest), all targets were linked against these libraries. This included
1847        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1848        libraries. This caused unwanted dependencies to be created. This problem
1849        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1850        pcretest is linked with readline.
1851    
1852    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1853        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1854        moved outside it again, because FALSE and TRUE are already defined in AIX,
1855        but BOOL is not.
1856    
1857    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1858        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1859    
1860    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1861        hyphens) following file names and line numbers when outputting matching
1862        lines. This is not true; no spaces are inserted. I have also clarified the
1863        wording for the --colour (or --color) option.
1864    
1865    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1866        was not coloured; this is different to GNU grep, so I have changed it to be
1867        the same.
1868    
1869    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1870        each matching line was coloured. Now it goes on to look for further matches
1871        of any of the test patterns, which is the same behaviour as GNU grep.
1872    
1873    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1874        doesn't make sense to accept an empty string match in pcregrep, so I have
1875        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1876        seems to be how GNU grep behaves. [But see later change 40 for release
1877        8.33.]
1878    
1879    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1880        start or after a newline", because the conditional assertion was not being
1881        correctly handled. The rule now is that both the assertion and what follows
1882        in the first alternative must satisfy the test.
1883    
1884    9.  If auto-callout was enabled in a pattern with a conditional group whose
1885        condition was an assertion, PCRE could crash during matching, both with
1886        pcre_exec() and pcre_dfa_exec().
1887    
1888    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1889        used for matching.
1890    
1891    11. Unicode property support in character classes was not working for
1892        characters (bytes) greater than 127 when not in UTF-8 mode.
1893    
1894    12. Added the -M command line option to pcretest.
1895    
1896    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1897    
1898    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1899    
1900    16. Added comments and documentation about mis-use of no_arg in the C++
1901        wrapper.
1902    
1903    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1904        from Martin Jerabek that uses macro names for all relevant character and
1905        string constants.
1906    
1907    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1908        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1909        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1910        these, but not everybody uses configure.
1911    
1912    19. A conditional group that had only one branch was not being correctly
1913        recognized as an item that could match an empty string. This meant that an
1914        enclosing group might also not be so recognized, causing infinite looping
1915        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1916        with the subject "ab", where knowledge that the repeated group can match
1917        nothing is needed in order to break the loop.
1918    
1919    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1920        exec(), but without supplying a callout function, matching went wrong.
1921    
1922    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1923        leak if the size of the offset vector was greater than 30. When the vector
1924        is smaller, the saved offsets during recursion go onto a local stack
1925        vector, but for larger vectors malloc() is used. It was failing to free
1926        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1927        error, in fact).
1928    
1929    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1930        heapframe that is used only when UTF-8 support is enabled. This caused no
1931        problem, but was untidy.
1932    
1933    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1934        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1935        included within another project.
1936    
1937    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1938        slightly modified by me:
1939    
1940          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1941              not building pcregrep.
1942    
1943          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1944              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1945    
1946    25. Forward references, both numeric and by name, in patterns that made use of
1947        duplicate group numbers, could behave incorrectly or give incorrect errors,
1948        because when scanning forward to find the reference group, PCRE was not
1949        taking into account the duplicate group numbers. A pattern such as
1950        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1951    
1952    26. Changed a few more instances of "const unsigned char *" to USPTR, making
1953        the feature of a custom pointer more persuasive (as requested by a user).
1954    
1955    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
1956        pcretest.c, inside #ifndefs, because it seems they are sometimes already
1957        pre-defined.
1958    
1959    28. Added support for (*UTF8) at the start of a pattern.
1960    
1961    29. Arrange for flags added by the "release type" setting in CMake to be shown
1962        in the configuration summary.
1963    
1964    
1965    Version 7.8 05-Sep-08
1966    ---------------------
1967    
1968    1.  Replaced UCP searching code with optimized version as implemented for Ad
1969        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
1970        stage table and inline lookup instead of a function, giving speed ups of 2
1971        to 5 times on some simple patterns that I tested. Permission was given to
1972        distribute the MultiStage2.py script that generates the tables (it's not in
1973        the tarball, but is in the Subversion repository).
1974    
1975    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
1976        scripts.
1977    
1978    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
1979        a group with a zero qualifier. The result of the study could be incorrect,
1980        or the function might crash, depending on the pattern.
1981    
1982    4.  Caseless matching was not working for non-ASCII characters in back
1983        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
1984        It now works when Unicode Property Support is available.
1985    
1986    5.  In pcretest, an escape such as \x{de} in the data was always generating
1987        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
1988        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
1989        truncation.
1990    
1991    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
1992    
1993    7.  Added two (int) casts to pcregrep when printing the difference of two
1994        pointers, in case they are 64-bit values.
1995    
1996    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
1997        test 2 if it fails.
1998    
1999    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2000        and a #define of that name to empty if it is not externally set. This is to
2001        allow users of MSVC to set it if necessary.
2002    
2003    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2004        the convenience functions in the pcre_get.c source file.
2005    
2006    11. An option change at the start of a pattern that had top-level alternatives
2007        could cause overwriting and/or a crash. This command provoked a crash in
2008        some environments:
2009    
2010          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2011    
2012        This potential security problem was recorded as CVE-2008-2371.
2013    
2014    12. For a pattern where the match had to start at the beginning or immediately
2015        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2016        pcre_dfa_exec() could read past the end of the passed subject if there was
2017        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2018        pcretest so that it places the subject at the end of its malloc-ed buffer.
2019    
2020    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2021        exec() might read past the end of the data buffer in UTF-8 mode.
2022    
2023    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2024        the data contained the byte 0x85 as part of a UTF-8 character within its
2025        first line. This applied both to normal and DFA matching.
2026    
2027    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2028        /^[^d]*?$/8 failed to match "abc".
2029    
2030    16. Added a missing copyright notice to pcrecpp_internal.h.
2031    
2032    17. Make it more clear in the documentation that values returned from
2033        pcre_exec() in ovector are byte offsets, not character counts.
2034    
2035    18. Tidied a few places to stop certain compilers from issuing warnings.
2036    
2037    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2038        supplied by Stefan Weber. I made a further small update for 7.8 because
2039        there is a change of source arrangements: the pcre_searchfuncs.c module is
2040        replaced by pcre_ucd.c.
2041    
2042    
2043    Version 7.7 07-May-08
2044    ---------------------
2045    
2046    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2047        a string to a long long, pretend we don't even have a long long." This is
2048        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2049    
2050    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2051        pre-7.6 versions, which defined a global no_arg variable instead of putting
2052        it in the RE class. (See also #8 below.)
2053    
2054    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2055        Lopes.
2056    
2057    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2058    
2059        (1) The include/exclude patterns were being applied to the whole pathnames
2060            of files, instead of just to the final components.
2061    
2062        (2) If there was more than one level of directory, the subdirectories were
2063            skipped unless they satisfied the include/exclude conditions. This is
2064            inconsistent with GNU grep (and could even be seen as contrary to the
2065            pcregrep specification - which I improved to make it absolutely clear).
2066            The action now is always to scan all levels of directory, and just
2067            apply the include/exclude patterns to regular files.
2068    
2069    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2070        --exclude_dir in the tests to avoid scanning .svn directories.
2071    
2072    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2073        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2074        doesn't support NULs in patterns.
2075    
2076    7.  Added some missing "const"s to declarations of static tables in
2077        pcre_compile.c and pcre_dfa_exec.c.
2078    
2079    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2080        caused by fix #2  above. (Subsequently also a second patch to fix the
2081        first patch. And a third patch - this was a messy problem.)
2082    
2083    9.  Applied Craig's patch to remove the use of push_back().
2084    
2085    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2086        matching function regexec().
2087    
2088    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2089        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2090        references. PCRE supports relative numbers with this syntax (I don't think
2091        Oniguruma does).
2092    
2093    12. Previously, a group with a zero repeat such as (...){0} was completely
2094        omitted from the compiled regex. However, this means that if the group
2095        was called as a subroutine from elsewhere in the pattern, things went wrong
2096        (an internal error was given). Such groups are now left in the compiled
2097        pattern, with a new opcode that causes them to be skipped at execution
2098        time.
2099    
2100    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2101        to the way PCRE behaves:
2102    
2103        (a) A lone ] character is dis-allowed (Perl treats it as data).
2104    
2105        (b) A back reference to an unmatched subpattern matches an empty string
2106            (Perl fails the current match path).
2107    
2108        (c) A data ] in a character class must be notated as \] because if the
2109            first data character in a class is ], it defines an empty class. (In
2110            Perl it is not possible to have an empty class.) The empty class []
2111            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2112            The negative empty class [^] matches any one character, independently
2113            of the DOTALL setting.
2114    
2115    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2116        non-existent subpattern following a character class starting with ']' and
2117        containing () gave an internal compiling error instead of "reference to
2118        non-existent subpattern". Fortunately, when the pattern did exist, the
2119        compiled code was correct. (When scanning forwards to check for the
2120        existence of the subpattern, it was treating the data ']' as terminating
2121        the class, so got the count wrong. When actually compiling, the reference
2122        was subsequently set up correctly.)
2123    
2124    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2125        it was being rejected as not supported by pcre_dfa_exec(), even though
2126        other assertions are supported. I have made pcre_dfa_exec() support
2127        (*FAIL).
2128    
2129    16. The implementation of 13c above involved the invention of a new opcode,
2130        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2131        cannot be changed at match time, I realized I could make a small
2132        improvement to matching performance by compiling OP_ALLANY instead of
2133        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2134        on the OP_ANY path.
2135    
2136    17. Compiling pcretest on Windows with readline support failed without the
2137        following two fixes: (1) Make the unistd.h include conditional on
2138        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2139    
2140    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2141        ncurses library to be included for pcretest when ReadLine support is
2142        requested, but also to allow for it to be overridden. This patch came from
2143        Daniel Bergström.
2144    
2145    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2146        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2147        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2148        spotting this.
2149    
2150    
2151    Version 7.6 28-Jan-08
2152    ---------------------
2153    
2154    1.  A character class containing a very large number of characters with
2155        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2156        overflow.
2157    
2158    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2159        HAVE_LONG_LONG is not defined.
2160    
2161    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2162        bring them up to date and include new features. This patch includes:
2163    
2164        - Fixed PH's badly added libz and libbz2 support.
2165        - Fixed a problem with static linking.
2166        - Added pcredemo. [But later removed - see 7 below.]
2167        - Fixed dftables problem and added an option.
2168        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2169            HAVE_LONG_LONG.
2170        - Added readline support for pcretest.
2171        - Added an listing of the option settings after cmake has run.
2172    
2173    4.  A user submitted a patch to Makefile that makes it easy to create
2174        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2175        Makefile.am that cause it to include this special target, without
2176        affecting anything else. Note that the same mingw target plus all
2177        the other distribution libraries and programs are now supported
2178        when configuring with CMake (see 6 below) instead of with
2179        Configure/Make.
2180    
2181    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2182        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2183        exported in the Windows port". It has not yet been confirmed that the patch
2184        solves the problem, but it does no harm.
2185    
2186    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2187        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2188        with CMake, and also correct the comment about stack recursion.
2189    
2190    7.  Remove the automatic building of pcredemo from the ./configure system and
2191        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2192        of a program that users should build themselves after PCRE is installed, so
2193        building it automatically is not really right. What is more, it gave
2194        trouble in some build environments.
2195    
2196    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2197    
2198    
2199    Version 7.5 10-Jan-08
2200    ---------------------
2201    
2202    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2203        values in parens when parsing an RE using the C++ wrapper."
2204    
2205    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2206        Characters greater than 255 were excluded from the class instead of being
2207        included.
2208    
2209    3.  The same bug as (2) above applied to negated POSIX classes such as
2210        [:^space:].
2211    
2212    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2213        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2214        I have changed it.
2215    
2216    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2217        first named subpattern) and a construct such as (?&a) would reference the
2218        first named subpattern whose name started with "a" (in other words, the
2219        length check was missing). Both these problems are fixed. "Subpattern name
2220        expected" is now given for (?&) (a zero-length name), and this patch also
2221        makes it give the same error for \k'' (previously it complained that that
2222        was a reference to a non-existent subpattern).
2223    
2224    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2225        this is right because (?- can be followed by option settings as well as by
2226        digits. I have, however, made the messages clearer.
2227    
2228    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2229        than the number used in the conditional) now cause a compile-time error.
2230        This is actually not compatible with Perl, which accepts such patterns, but
2231        treats the conditional as always being FALSE (as PCRE used to), but it
2232        seems to me that giving a diagnostic is better.
2233    
2234    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2235        and messages.
2236    
2237    9.  Fix two occurrences of "backslash" in comments that should have been
2238        "backspace".
2239    
2240    10. Remove two redundant lines of code that can never be obeyed (their function
2241        was moved elsewhere).
2242    
2243    11. The program that makes PCRE's Unicode character property table had a bug
2244        which caused it to generate incorrect table entries for sequences of
2245        characters that have the same character type, but are in different scripts.
2246        It amalgamated them into a single range, with the script of the first of
2247        them. In other words, some characters were in the wrong script. There were
2248        thirteen such cases, affecting characters in the following ranges:
2249    
2250          U+002b0 - U+002c1
2251          U+0060c - U+0060d
2252          U+0061e - U+00612
2253          U+0064b - U+0065e
2254          U+0074d - U+0076d
2255          U+01800 - U+01805
2256          U+01d00 - U+01d77
2257          U+01d9b - U+01dbf
2258          U+0200b - U+0200f
2259          U+030fc - U+030fe
2260          U+03260 - U+0327f
2261          U+0fb46 - U+0fbb1
2262          U+10450 - U+1049d
2263    
2264    12. The -o option (show only the matching part of a line) for pcregrep was not
2265        compatible with GNU grep in that, if there was more than one match in a
2266        line, it showed only the first of them. It now behaves in the same way as
2267        GNU grep.
2268    
2269    13. If the -o and -v options were combined for pcregrep, it printed a blank
2270        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2271        does the same. The return code can be used to tell if there were any
2272        non-matching lines.
2273    
2274    14. Added --file-offsets and --line-offsets to pcregrep.
2275    
2276    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2277        infinitely looping recursion. The bug was that positive lookaheads were not
2278        being skipped when checking for a possible empty match (negative lookaheads
2279        and both kinds of lookbehind were skipped).
2280    
2281    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2282        inclusion of <windows.h> to before rather than after the definition of
2283        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2284    
2285    17. Specifying a possessive quantifier with a specific limit for a Unicode
2286        character property caused pcre_compile() to compile bad code, which led at
2287        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2288        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2289        caused the error; without that there was no problem.
2290    
2291    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2292    
2293    19. Added --enable-pcretest-libreadline.
2294    
2295    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2296        RE::GlobalReplace(). As a result, the number of replacements returned was
2297        double what it should be. I removed one of the increments, but Craig sent a
2298        later patch that removed the other one (the right fix) and added unit tests
2299        that check the return values (which was not done before).
2300    
2301    21. Several CMake things:
2302    
2303        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2304            the names libpcre and libpcreposix, not just pcre and pcreposix.
2305    
2306        (2) The above change means that pcretest and pcregrep are now correctly
2307            linked with the newly-built libraries, not previously installed ones.
2308    
2309        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2310    
2311    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2312        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2313        UTF-8 newline character). The key issue is that the pattern starts .*;
2314        this means that the match must be either at the beginning, or after a
2315        newline. The bug was in the code for advancing after a failed match and
2316        checking that the new position followed a newline. It was not taking
2317        account of UTF-8 characters correctly.
2318    
2319    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2320        character classes. PCRE was not treating the sequence [:...:] as a
2321        character class unless the ... were all letters. Perl, however, seems to
2322        allow any characters between [: and :], though of course it rejects as
2323        unknown any "names" that contain non-letters, because all the known class
2324        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2325        for example, whereas PCRE did not - it did not recognize a POSIX character
2326        class. This seemed a bit dangerous, so the code has been changed to be
2327        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2328        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2329        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2330        Perl does, and where it didn't before.
2331    
2332    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2333        Windows environments %n is disabled by default.
2334    
2335    
2336    Version 7.4 21-Sep-07
2337    ---------------------
2338    
2339    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2340        means that a class such as [\s] counted as "explicit reference to CR or
2341        LF". That isn't really right - the whole point of the change was to try to
2342        help when there was an actual mention of one of the two characters. So now
2343        the change happens only if \r or \n (or a literal CR or LF) character is
2344        encountered.
2345    
2346    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2347        of both had grown to the point where there were only 3 bits left.
2348        Fortunately, there was spare space in the data structure, and so I have
2349        moved the internal flags into a new 16-bit field to free up more option
2350        bits.
2351    
2352    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2353        but did not set the internal JCHANGED flag - either of these is enough to
2354        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2355        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2356        start sets both bits.
2357    
2358    4.  Added options (at build time, compile time, exec time) to change \R from
2359        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2360    
2361    5.  doc/pcresyntax.html was missing from the distribution.
2362    
2363    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2364        compatibility, even though it is no longer used.
2365    
2366    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2367        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2368        windows.h file is present (where different names are used). [This was
2369        reversed later after testing - see 16 below.]
2370    
2371    8.  Changed all #include <config.h> to #include "config.h". There were also
2372        some further <pcre.h> cases that I changed to "pcre.h".
2373    
2374    9.  When pcregrep was used with the --colour option, it missed the line ending
2375        sequence off the lines that it output.
2376    
2377    10. It was pointed out to me that arrays of string pointers cause lots of
2378        relocations when a shared library is dynamically loaded. A technique of
2379        using a single long string with a table of offsets can drastically reduce
2380        these. I have refactored PCRE in four places to do this. The result is
2381        dramatic:
2382    
2383          Originally:                          290
2384          After changing UCP table:            187
2385          After changing error message table:   43
2386          After changing table of "verbs"       36
2387          After changing table of Posix names   22
2388    
2389        Thanks to the folks working on Gregex for glib for this insight.
2390    
2391    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2392        unicode-properties was also set.
2393    
2394    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2395    
2396    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2397        checked only for CRLF.
2398    
2399    14. Added casts to pcretest.c to avoid compiler warnings.
2400    
2401    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2402    
2403    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2404        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2405        entirely. This removes changes made in 7 above.
2406    
2407    17. The CMake files have been updated, and there is now more information about
2408        building with CMake in the NON-UNIX-USE document.
2409    
2410    
2411    Version 7.3 28-Aug-07
2412    ---------------------
2413    
2414     1. In the rejigging of the build system that eventually resulted in 7.1, the
2415        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2416        brackets there is not right, since it causes compilers to look for an
2417        installed pcre.h, not the version that is in the source that is being
2418        compiled (which of course may be different). I have changed it back to:
2419    
2420          #include "pcre.h"
2421    
2422        I have a vague recollection that the change was concerned with compiling in
2423        different directories, but in the new build system, that is taken care of
2424        by the VPATH setting the Makefile.
2425    
2426     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2427        when the subject happened to end in the byte 0x85 (e.g. if the last
2428        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2429        characters but of course it shouldn't be taken as a newline when it is part
2430        of another character. The bug was that, for an unlimited repeat of . in
2431        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2432        characters when looking for a newline.
2433    
2434     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2435    
2436     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2437        in debug output.
2438    
2439     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2440        long printing in the pcrecpp unittest when running under MinGW.
2441    
2442     6. ESC_K was left out of the EBCDIC table.
2443    
2444     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2445        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2446        limit also applies to "virtual nesting" when a pattern is recursive, and in
2447        this case 1000 isn't so big. I have been able to remove this limit at the
2448        expense of backing off one optimization in certain circumstances. Normally,
2449        when pcre_exec() would call its internal match() function recursively and
2450        immediately return the result unconditionally, it uses a "tail recursion"
2451        feature to save stack. However, when a subpattern that can match an empty
2452        string has an unlimited repetition quantifier, it no longer makes this
2453        optimization. That gives it a stack frame in which to save the data for
2454        checking that an empty string has been matched. Previously this was taken
2455        from the 1000-entry workspace that had been reserved. So now there is no
2456        explicit limit, but more stack is used.
2457    
2458     8. Applied Daniel's patches to solve problems with the import/export magic
2459        syntax that is required for Windows, and which was going wrong for the
2460        pcreposix and pcrecpp parts of the library. These were overlooked when this
2461        problem was solved for the main library.
2462    
2463     9. There were some crude static tests to avoid integer overflow when computing
2464        the size of patterns that contain repeated groups with explicit upper
2465        limits. As the maximum quantifier is 65535, the maximum group length was
2466        set at 30,000 so that the product of these two numbers did not overflow a
2467        32-bit integer. However, it turns out that people want to use groups that
2468        are longer than 30,000 bytes (though not repeat them that many times).
2469        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2470        made it possible to implement the integer overflow checks in a much more
2471        dynamic way, which I have now done. The artificial limitation on group
2472        length has been removed - we now have only the limit on the total length of
2473        the compiled pattern, which depends on the LINK_SIZE setting.
2474    
2475    10. Fixed a bug in the documentation for get/copy named substring when
2476        duplicate names are permitted. If none of the named substrings are set, the
2477        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2478        empty string.
2479    
2480    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2481        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2482        because the ] is interpreted as the first data character and the
2483        terminating ] is not found. PCRE has been made compatible with Perl in this
2484        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2485        cause memory overwriting.
2486    
2487    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2488        string has been matched (to stop an infinite loop). It was not recognizing
2489        a conditional subpattern that could match an empty string if that
2490        subpattern was within another subpattern. For example, it looped when
2491        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2492        condition was not nested. This bug has been fixed.
2493    
2494    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2495        past the start of the subject in the presence of bytes with the top bit
2496        set, for example "\x8aBCD".
2497    
2498    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2499        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2500    
2501    14. Optimized (?!) to (*FAIL).
2502    
2503    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2504        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2505        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2506        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2507        does: it's just the validity check that is more restrictive.
2508    
2509    16. Inserted checks for integer overflows during escape sequence (backslash)
2510        processing, and also fixed erroneous offset values for syntax errors during
2511        backslash processing.
2512    
2513    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2514        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2515    
2516    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2517        caused an overrun.
2518    
2519    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2520        something other than just ASCII characters) inside a group that had an
2521        unlimited repeat caused a loop at compile time (while checking to see
2522        whether the group could match an empty string).
2523    
2524    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2525        [\P{Any}] did so. (Error in the code for printing property names.)
2526    
2527    21. An orphan \E inside a character class could cause a crash.
2528    
2529    22. A repeated capturing bracket such as (A)? could cause a wild memory
2530        reference during compilation.
2531    
2532    23. There are several functions in pcre_compile() that scan along a compiled
2533        expression for various reasons (e.g. to see if it's fixed length for look
2534        behind). There were bugs in these functions when a repeated \p or \P was
2535        present in the pattern. These operators have additional parameters compared
2536        with \d, etc, and these were not being taken into account when moving along
2537        the compiled data. Specifically:
2538    
2539        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2540            length.
2541    
2542        (b) An item such as \pL+ within a repeated group could cause crashes or
2543            loops.
2544    
2545        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2546            "reference to non-existent subpattern" error.
2547    
2548        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2549    
2550    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2551        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2552    
2553    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2554    
2555    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2556        character were causing crashes (broken optimization).
2557    
2558    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2559        \p or \P) caused a compile-time loop.
2560    
2561    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2562        break. For example, the unstudied pattern [\r\n]A does not match the string
2563        "\r\nA" because change 7.0/46 below moves the current point on by two
2564        characters after failing to match at the start. However, the pattern \nA
2565        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2566        the same is true. There doesn't seem any very clean way out of this, but
2567        what I have chosen to do makes the common cases work: PCRE now takes note
2568        of whether there can be an explicit match for \r or \n anywhere in the
2569        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2570        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2571        pattern has explicit CR or LF references.
2572    
2573    29. Added (*CR) etc for changing newline setting at start of pattern.
2574    
2575    
2576    Version 7.2 19-Jun-07
2577  ---------------------  ---------------------
2578    
2579   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,   1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
# Line 21  Version 7.2 01-May-07 Line 2593  Version 7.2 01-May-07
2593      stack recursion. This gives a massive performance boost under BSD, but just      stack recursion. This gives a massive performance boost under BSD, but just
2594      a small improvement under Linux. However, it saves one field in the frame      a small improvement under Linux. However, it saves one field in the frame
2595      in all cases.      in all cases.
2596    
2597   6. Added more features from the forthcoming Perl 5.10:   6. Added more features from the forthcoming Perl 5.10:
2598    
2599      (a) (?-n) (where n is a string of digits) is a relative subroutine or      (a) (?-n) (where n is a string of digits) is a relative subroutine or
2600          recursion call. It refers to the nth most recently opened parentheses.          recursion call. It refers to the nth most recently opened parentheses.
2601    
2602      (b) (?+n) is also a relative subroutine call; it refers to the nth next      (b) (?+n) is also a relative subroutine call; it refers to the nth next
2603          to be opened parentheses.          to be opened parentheses.
2604    
2605      (c) Conditions that refer to capturing parentheses can be specified      (c) Conditions that refer to capturing parentheses can be specified
2606          relatively, for example, (?(-2)... or (?(+3)...          relatively, for example, (?(-2)... or (?(+3)...
2607    
2608        (d) \K resets the start of the current match so that everything before
2609            is not part of it.
2610    
2611        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2612    
2613        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2614            reference syntax.
2615    
2616        (g) (?| introduces a group in which the numbering of parentheses in each
2617            alternative starts with the same number.
2618    
2619        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2620    
2621     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2622        PCRE_INFO_JCHANGED.
2623    
2624     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2625        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2626        for detecting groups that can match an empty string.
2627    
2628     9. A pattern with a very large number of alternatives (more than several
2629        hundred) was running out of internal workspace during the pre-compile
2630        phase, where pcre_compile() figures out how much memory will be needed. A
2631        bit of new cunning has reduced the workspace needed for groups with
2632        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2633        workspace instead of running out of the 4096 that are available.
2634    
2635    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2636    
2637    11. Applied patch from Google to remove an optimization that didn't quite work.
2638        The report of the bug said:
2639    
2640          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2641          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2642          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2643    
2644    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2645        it matched the wrong number of bytes.
2646    
2647    
2648  Version 7.1 24-Apr-07  Version 7.1 24-Apr-07
2649  ---------------------  ---------------------

Legend:
Removed from v.167  
changed lines
  Added in v.1407

  ViewVC Help
Powered by ViewVC 1.1.5