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

Diff of /code/trunk/ChangeLog

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

revision 127 by ph10, Mon Mar 19 11:44:45 2007 UTC revision 1450 by ph10, Fri Jan 17 17:50:51 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
3    
4  Version 7.1 12-Mar-07  Version 8.35-RC1 xx-xxxx-201x
5    -----------------------------
6    
7    1.  A new flag is set, when property checks are present in an XCLASS.
8        When this flag is not set, PCRE can perform certain optimizations
9        such as studying these XCLASS-es.
10    
11    2.  The auto-possessification of character sets were improved: a normal
12        and an extended character set can be compared now. Furthermore
13        the JIT compiler optimizes more character set checks.
14    
15    3.  Got rid of some compiler warnings for potentially uninitialized variables
16        that show up only when compiled with -O2.
17    
18    4.  A pattern such as (?=ab\K) that uses \K in an assertion can set the start
19        of a match later then the end of the match. The pcretest program was not
20        handling the case sensibly - it was outputting from the start to the next
21        binary zero. It now reports this situation in a message, and outputs the
22        text from the end to the start.
23    
24    5.  Fast forward search is improved in JIT. Instead of the first three
25        characters, any three characters with fixed position can be searched.
26        Search order: first, last, middle.
27    
28    6.  Improve character range checks in JIT. Characters are read by an inprecise
29        function now, which returns with an unknown value if the character code is
30        above a certain treshold (e.g: 256). The only limitation is that the value
31        must be bigger than the treshold as well. This function is useful, when
32        the characters above the treshold are handled in the same way.
33    
34    7.  The macros whose names start with RAWUCHAR are placeholders for a future
35        mode in which only the bottom 21 bits of 32-bit data items are used. To
36        make this more memorable for those maintaining the code, the names have
37        been changed to start with UCHAR21, and an extensive comment has been added
38        to their definition.
39    
40    8.  Add missing (new) files sljitNativeTILEGX.c and sljitNativeTILEGX-encoder.c
41        to the export list in Makefile.am (they were accidentally omitted from the
42        8.34 tarball).
43    
44    9.  The informational output from pcretest used the phrase "starting byte set"
45        which is inappropriate for the 16-bit and 32-bit libraries. As the output
46        for "first char" and "need char" really means "non-UTF-char", I've changed
47        "byte" to "char", and slightly reworded the output. The documentation about
48        these values has also been (I hope) clarified.
49    
50    10. Another JIT related optimization: use table jumps for selecting the correct
51        backtracking path, when more than four alternatives are present inside a
52        bracket.
53    
54    11. Empty match is not possible, when the minimum length is greater than zero,
55        and there is no \K in the pattern. JIT should avoid empty match checks in
56        such cases.
57    
58    12. In a caseless character class with UCP support, when a character with more
59        than one alternative case was not the first character of a range, not all
60        the alternative cases were added to the class. For example, s and \x{17f}
61        are both alternative cases for S: the class [RST] was handled correctly,
62        but [R-T] was not.
63    
64    13. The configure.ac file always checked for pthread support when JIT was
65        enabled. This is not used in Windows, so I have put this test inside a
66        check for the presence of windows.h (which was already tested for).
67    
68    14. Improve pattern prefix search by a simplified Boyer-Moore algorithm in JIT.
69        The algorithm provides a way to skip certain starting offsets, and usually
70        faster than linear prefix searches.
71    
72    15. Change 13 for 8.20 updated RunTest to check for the 'fr' locale as well
73        as for 'fr_FR' and 'french'. For some reason, however, it then used the
74        Windows-specific input and output files, which have 'french' screwed in.
75        So this could never have worked. One of the problems with locales is that
76        they aren't always the same. I have now updated RunTest so that it checks
77        the output of the locale test (test 3) against three different output
78        files, and it allows the test to pass if any one of them matches. With luck
79        this should make the test pass on some versions of Solaris where it was
80        failing. Because of the uncertainty, the script did not used to stop if
81        test 3 failed; it now does. If further versions of a French locale ever
82        come to light, they can now easily be added.
83    
84    16. If --with-pcregrep-bufsize was given a non-integer value such as "50K",
85        there was a message during ./configure, but it did not stop. This now
86        provokes an error. The invalid example in README has been corrected.
87        If a value less than the minimum is given, the minimum value has always
88        been used, but now a warning is given.
89    
90    17. If --enable-bsr-anycrlf was set, the special 16/32-bit test failed. This
91        was a bug in the test system, which is now fixed. Also, the list of various
92        configurations that are tested for each release did not have one with both
93        16/32 bits and --enable-bar-anycrlf. It now does.
94    
95    18. pcretest was missing "-C bsr" for displaying the \R default setting.
96    
97    
98    Version 8.34 15-December-2013
99    -----------------------------
100    
101    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
102        executable memory. Patch inspired by Carsten Klein.
103    
104    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
105        this macro is never tested and has no effect, because the work to support
106        coverage involves only compiling and linking options and special targets in
107        the Makefile. The comment in config.h implied that defining the macro would
108        enable coverage support, which is totally false. There was also support for
109        setting this macro in the CMake files (my fault, I just copied it from
110        configure). SUPPORT_GCOV has now been removed.
111    
112    3.  Make a small performance improvement in strlen16() and strlen32() in
113        pcretest.
114    
115    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
116        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
117        these cases). There was also one in pcretest.c.
118    
119    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
120    
121    6.  In UTF mode, the code for checking whether a group could match an empty
122        string (which is used for indefinitely repeated groups to allow for
123        breaking an infinite loop) was broken when the group contained a repeated
124        negated single-character class with a character that occupied more than one
125        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
126        UTF-8 mode). The effect was undefined: the group might or might not be
127        deemed as matching an empty string, or the program might have crashed.
128    
129    7.  The code for checking whether a group could match an empty string was not
130        recognizing that \h, \H, \v, \V, and \R must match a character.
131    
132    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
133        an empty string. If it can, pcretest shows this in its information output.
134    
135    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
136        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
137        matched by pcre_exec() without using JIT:
138    
139        (a) If the rest of the pattern did not match after a maximal run of
140            grapheme clusters, the code for backing up to try with fewer of them
141            did not always back up over a full grapheme when characters that do not
142            have the modifier quality were involved, e.g. Hangul syllables.
143    
144        (b) If the match point in a subject started with modifier character, and
145            there was no match, the code could incorrectly back up beyond the match
146            point, and potentially beyond the first character in the subject,
147            leading to a segfault or an incorrect match result.
148    
149    10. A conditional group with an assertion condition could lead to PCRE
150        recording an incorrect first data item for a match if no other first data
151        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
152        first data item, and therefore matched "ca" after "c" instead of at the
153        start.
154    
155    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
156        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
157    
158    12. The source of pcregrep now includes z/OS-specific code so that it can be
159        compiled for z/OS as part of the special z/OS distribution.
160    
161    13. Added the -T and -TM options to pcretest.
162    
163    14. The code in pcre_compile.c for creating the table of named capturing groups
164        has been refactored. Instead of creating the table dynamically during the
165        actual compiling pass, the information is remembered during the pre-compile
166        pass (on the stack unless there are more than 20 named groups, in which
167        case malloc() is used) and the whole table is created before the actual
168        compile happens. This has simplified the code (it is now nearly 150 lines
169        shorter) and prepared the way for better handling of references to groups
170        with duplicate names.
171    
172    15. A back reference to a named subpattern when there is more than one of the
173        same name now checks them in the order in which they appear in the pattern.
174        The first one that is set is used for the reference. Previously only the
175        first one was inspected. This change makes PCRE more compatible with Perl.
176    
177    16. Unicode character properties were updated from Unicode 6.3.0.
178    
179    17. The compile-time code for auto-possessification has been refactored, based
180        on a patch by Zoltan Herczeg. It now happens after instead of during
181        compilation. The code is cleaner, and more cases are handled. The option
182        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
183        options in pcretest are provided to set it. It can also be set by
184        (*NO_AUTO_POSSESS) at the start of a pattern.
185    
186    18. The character VT has been added to the default ("C" locale) set of
187        characters that match \s and are generally treated as white space,
188        following this same change in Perl 5.18. There is now no difference between
189        "Perl space" and "POSIX space". Whether VT is treated as white space in
190        other locales depends on the locale.
191    
192    19. The code for checking named groups as conditions, either for being set or
193        for being recursed, has been refactored (this is related to 14 and 15
194        above). Processing unduplicated named groups should now be as fast at
195        numerical groups, and processing duplicated groups should be faster than
196        before.
197    
198    20. Two patches to the CMake build system, by Alexander Barkov:
199    
200          (1) Replace the "source" command by "." in CMakeLists.txt because
201              "source" is a bash-ism.
202    
203          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
204              without these the CMake build does not work on Solaris.
205    
206    21. Perl has changed its handling of \8 and \9. If there is no previously
207        encountered capturing group of those numbers, they are treated as the
208        literal characters 8 and 9 instead of a binary zero followed by the
209        literals. PCRE now does the same.
210    
211    22. Following Perl, added \o{} to specify codepoints in octal, making it
212        possible to specify values greater than 0777 and also making them
213        unambiguous.
214    
215    23. Perl now gives an error for missing closing braces after \x{... instead of
216        treating the string as literal. PCRE now does the same.
217    
218    24. RunTest used to grumble if an inappropriate test was selected explicitly,
219        but just skip it when running all tests. This make it awkward to run ranges
220        of tests when one of them was inappropriate. Now it just skips any
221        inappropriate tests, as it always did when running all tests.
222    
223    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
224        character types such as \d or \w, too many callouts were inserted, and the
225        data that they returned was rubbish.
226    
227    26. In UCP mode, \s was not matching two of the characters that Perl matches,
228        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
229        were matched by \h. The code has now been refactored so that the lists of
230        the horizontal and vertical whitespace characters used for \h and \v (which
231        are defined only in one place) are now also used for \s.
232    
233    27. Add JIT support for the 64 bit TileGX architecture.
234        Patch by Jiong Wang (Tilera Corporation).
235    
236    28. Possessive quantifiers for classes (both explicit and automatically
237        generated) now use special opcodes instead of wrapping in ONCE brackets.
238    
239    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
240        (because it's meaningless), this was not happening when PCRE_CASELESS was
241        set. Not wrong, but inefficient.
242    
243    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
244        properties for \w, \d, etc) is present in a test regex. Otherwise if the
245        test contains no characters greater than 255, Perl doesn't realise it
246        should be using Unicode semantics.
247    
248    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
249        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
250        does in Unicode mode.
251    
252    32. Added the "forbid" facility to pcretest so that putting tests into the
253        wrong test files can sometimes be quickly detected.
254    
255    33. There is now a limit (default 250) on the depth of nesting of parentheses.
256        This limit is imposed to control the amount of system stack used at compile
257        time. It can be changed at build time by --with-parens-nest-limit=xxx or
258        the equivalent in CMake.
259    
260    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
261        errors. Perl warns for these when in warning mode, but PCRE has no facility
262        for giving warnings.
263    
264    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
265        However, this was not working for (?!) because it is optimized to (*FAIL),
266        for which PCRE does not allow quantifiers. The optimization is now disabled
267        when a quantifier follows (?!). I can't see any use for this, but it makes
268        things uniform.
269    
270    36. Perl no longer allows group names to start with digits, so I have made this
271        change also in PCRE. It simplifies the code a bit.
272    
273    37. In extended mode, Perl ignores spaces before a + that indicates a
274        possessive quantifier. PCRE allowed a space before the quantifier, but not
275        before the possessive +. It now does.
276    
277    38. The use of \K (reset reported match start) within a repeated possessive
278        group such as (a\Kb)*+ was not working.
279    
280    40. Document that the same character tables must be used at compile time and
281        run time, and that the facility to pass tables to pcre_exec() and
282        pcre_dfa_exec() is for use only with saved/restored patterns.
283    
284    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
285        features for Builds with MSVC:
286    
287        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
288           builds). The .rc files can be used to set FileDescription and many other
289           attributes.
290    
291        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
292           This allows higher-level build scripts which want .pdb files to avoid
293           hard-coding the exact files needed."
294    
295    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
296        mean "start of word" and "end of word", respectively, as a transition aid.
297    
298    43. A minimizing repeat of a class containing codepoints greater than 255 in
299        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
300        compiled to use the heap for recursion.
301    
302    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
303        is configured.
304    
305    
306    Version 8.33 28-May-2013
307    ------------------------
308    
309    1.  Added 'U' to some constants that are compared to unsigned integers, to
310        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
311        variables that are added to signed variables, to ensure the result is
312        signed and can be negated.
313    
314    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
315        CMake config files.
316    
317    3.  Revise the creation of config.h.generic so that all boolean macros are
318        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
319        overriding via -D on the command line possible.
320    
321    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
322        to unsigned int is reported to make a quite noticeable speed difference in
323        a specific Windows environment. Testing on Linux did also appear to show
324        some benefit (and it is clearly not harmful). Also fixed the definition of
325        Xop which should be unsigned.
326    
327    5.  Related to (4), changing the definition of the intermediate variable cc
328        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
329        improvements.
330    
331    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
332        unnecessary spaces.
333    
334    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
335        and later.
336    
337    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
338    
339    9.  Optimizing fast_forward_start_bits in JIT.
340    
341    10. Adding support for callouts in JIT, and fixing some issues revealed
342        during this work. Namely:
343    
344        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
345    
346        (b) Minimum length was not checked before the matching is started.
347    
348    11. The value of capture_last that is passed to callouts was incorrect in some
349        cases when there was a capture on one path that was subsequently abandoned
350        after a backtrack. Also, the capture_last value is now reset after a
351        recursion, since all captures are also reset in this case.
352    
353    12. The interpreter no longer returns the "too many substrings" error in the
354        case when an overflowing capture is in a branch that is subsequently
355        abandoned after a backtrack.
356    
357    13. In the pathological case when an offset vector of size 2 is used, pcretest
358        now prints out the matched string after a yield of 0 or 1.
359    
360    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
361        Only supported by the JIT compiler at the moment.
362    
363    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
364    
365    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
366        offset of the starting point of the matching process, provided the offsets
367        vector is large enough.
368    
369    17. The \A escape now records a lookbehind value of 1, though its execution
370        does not actually inspect the previous character. This is to ensure that,
371        in partial multi-segment matching, at least one character from the old
372        segment is retained when a new segment is processed. Otherwise, if there
373        are no lookbehinds in the pattern, \A might match incorrectly at the start
374        of a new segment.
375    
376    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
377    
378    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
379        gives some modest performance improvement in 8-bit mode.
380    
381    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
382        be expressed in certain programming languages using Universal Character
383        Names.
384    
385    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
386        which points out that "non characters" are not "characters that may not
387        appear in Unicode strings" but rather "characters that are reserved for
388        internal use and have only local meaning".
389    
390    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
391        there was a conditional group that depended on an assertion, if the
392        assertion was false, the callout that immediately followed the alternation
393        in the condition was skipped when pcre_exec() was used for matching.
394    
395    23. Allow an explicit callout to be inserted before an assertion that is the
396        condition for a conditional group, for compatibility with automatic
397        callouts, which always insert a callout at this point.
398    
399    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
400        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
401    
402    25. (*PRUNE) is now supported by the JIT compiler.
403    
404    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
405    
406    27. Fix the case where there are two or more SKIPs with arguments that may be
407        ignored.
408    
409    28. (*SKIP) is now supported by the JIT compiler.
410    
411    29. (*THEN) is now supported by the JIT compiler.
412    
413    30. Update RunTest with additional test selector options.
414    
415    31. The way PCRE handles backtracking verbs has been changed in two ways.
416    
417        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
418        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
419        backtracking. In some cases this makes it more Perl-compatible, but Perl's
420        rather obscure rules do not always do the same thing.
421    
422        (2) Previously, backtracking verbs were confined within assertions. This is
423        no longer the case for positive assertions, except for (*ACCEPT). Again,
424        this sometimes improves Perl compatibility, and sometimes does not.
425    
426    32. A number of tests that were in test 2 because Perl did things differently
427        have been moved to test 1, because either Perl or PCRE has changed, and
428        these tests are now compatible.
429    
430    32. Backtracking control verbs are now handled in the same way in JIT and
431        interpreter.
432    
433    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
434        contained a forward subroutine reference caused a compile error.
435    
436    34. Auto-detect and optimize limited repetitions in JIT.
437    
438    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
439        blocking (*UTF) etc.
440    
441    36. In the interpreter, maximizing pattern repetitions for characters and
442        character types now use tail recursion, which reduces stack usage.
443    
444    37. The value of the max lookbehind was not correctly preserved if a compiled
445        and saved regex was reloaded on a host of different endianness.
446    
447    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
448        of the compiled pattern block, expand the flags field from 16 to 32 bits
449        because it was almost full.
450    
451    39. Try madvise first before posix_madvise.
452    
453    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
454        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
455        original change locked out all matches of empty strings. This has been
456        changed so that one match of an empty string per line is recognized.
457        Subsequent searches on the same line (for colouring or for --only-matching,
458        for example) do not recognize empty strings.
459    
460    41. Applied a user patch to fix a number of spelling mistakes in comments.
461    
462    42. Data lines longer than 65536 caused pcretest to crash.
463    
464    43. Clarified the data type for length and startoffset arguments for pcre_exec
465        and pcre_dfa_exec in the function-specific man pages, where they were
466        explicitly stated to be in bytes, never having been updated. I also added
467        some clarification to the pcreapi man page.
468    
469    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
470        a segmentation fault.
471    
472    
473    Version 8.32 30-November-2012
474    -----------------------------
475    
476    1.  Improved JIT compiler optimizations for first character search and single
477        character iterators.
478    
479    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
480        Patch by Daniel Richard G.
481    
482    3.  Single character iterator optimizations in the JIT compiler.
483    
484    4.  Improved JIT compiler optimizations for character ranges.
485    
486    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
487        Reported by Giuseppe D'Angelo.
488    
489    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
490        of a line, was being set incorrectly in cases where .* appeared inside
491        atomic brackets at the start of a pattern, or where there was a subsequent
492        *PRUNE or *SKIP.
493    
494    7.  Improved instruction cache flush for POWER/PowerPC.
495        Patch by Daniel Richard G.
496    
497    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
498        grep:
499    
500        (a) There is now no limit to the number of patterns to be matched.
501    
502        (b) An error is given if a pattern is too long.
503    
504        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
505            are now supported.
506    
507        (d) --exclude-from and --include-from (multiple use) have been added.
508    
509        (e) Exclusions and inclusions now apply to all files and directories, not
510            just to those obtained from scanning a directory recursively.
511    
512        (f) Multiple uses of -f and --file-list are now supported.
513    
514        (g) In a Windows environment, the default for -d has been changed from
515            "read" (the GNU grep default) to "skip", because otherwise the presence
516            of a directory in the file list provokes an error.
517    
518        (h) The documentation has been revised and clarified in places.
519    
520    9.  Improve the matching speed of capturing brackets.
521    
522    10. Changed the meaning of \X so that it now matches a Unicode extended
523        grapheme cluster.
524    
525    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
526        out POSIX threads when JIT support is configured.
527    
528    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
529    
530    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
531        structure could go wrong in environments where size_t is not the same size
532        as int.
533    
534    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
535        set.
536    
537    15. The EBCDIC support had decayed; later updates to the code had included
538        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
539        general tidy up of EBCDIC-related issues, and the documentation was also
540        not quite right. There is now a test that can be run on ASCII systems to
541        check some of the EBCDIC-related things (but is it not a full test).
542    
543    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
544        in a small tidy to the code.
545    
546    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
547    
548    18. If the --only-matching (-o) option in pcregrep is specified multiple
549        times, each one causes appropriate output. For example, -o1 -o2 outputs the
550        substrings matched by the 1st and 2nd capturing parentheses. A separating
551        string can be specified by --om-separator (default empty).
552    
553    19. Improving the first n character searches.
554    
555    20. Turn case lists for horizontal and vertical white space into macros so that
556        they are defined only once.
557    
558    21. This set of changes together give more compatible Unicode case-folding
559        behaviour for characters that have more than one other case when UCP
560        support is available.
561    
562        (a) The Unicode property table now has offsets into a new table of sets of
563            three or more characters that are case-equivalent. The MultiStage2.py
564            script that generates these tables (the pcre_ucd.c file) now scans
565            CaseFolding.txt instead of UnicodeData.txt for character case
566            information.
567    
568        (b) The code for adding characters or ranges of characters to a character
569            class has been abstracted into a generalized function that also handles
570            case-independence. In UTF-mode with UCP support, this uses the new data
571            to handle characters with more than one other case.
572    
573        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
574            whose other case is greater than 256 are now correctly matched
575            caselessly. Previously, the high codepoint matched the low one, but not
576            vice versa.
577    
578        (d) The processing of \h, \H, \v, and \ in character classes now makes use
579            of the new class addition function, using character lists defined as
580            macros alongside the case definitions of 20 above.
581    
582        (e) Caseless back references now work with characters that have more than
583            one other case.
584    
585        (f) General caseless matching of characters with more than one other case
586            is supported.
587    
588    22. Unicode character properties were updated from Unicode 6.2.0
589    
590    23. Improved CMake support under Windows. Patch by Daniel Richard G.
591    
592    24. Add support for 32-bit character strings, and UTF-32
593    
594    25. Major JIT compiler update (code refactoring and bugfixing).
595        Experimental Sparc 32 support is added.
596    
597    26. Applied a modified version of Daniel Richard G's patch to create
598        pcre.h.generic and config.h.generic by "make" instead of in the
599        PrepareRelease script.
600    
601    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
602        pcre_compile.c when checking for a zero character.
603    
604    28. Introducing a native interface for JIT. Through this interface, the compiled
605        machine code can be directly executed. The purpose of this interface is to
606        provide fast pattern matching, so several sanity checks are not performed.
607        However, feature tests are still performed. The new interface provides
608        1.4x speedup compared to the old one.
609    
610    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
611        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
612        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
613    
614    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
615        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
616        the "old" RFC 2279). Instead, it ended up passing a negative length to
617        pcre_exec().
618    
619    31. Add support for GCC's visibility feature to hide internal functions.
620    
621    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
622        "unknown -C option" after outputting 0 or 1.
623    
624    33. There is now support for generating a code coverage report for the test
625        suite in environments where gcc is the compiler and lcov is installed. This
626        is mainly for the benefit of the developers.
627    
628    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
629        unaddressable using valgrind annotations, allowing valgrind to detect
630        invalid memory accesses. This is mainly for the benefit of the developers.
631    
632    25. (*UTF) can now be used to start a pattern in any of the three libraries.
633    
634    26. Give configure error if --enable-cpp but no C++ compiler found.
635    
636    
637    Version 8.31 06-July-2012
638    -------------------------
639    
640    1.  Fixing a wrong JIT test case and some compiler warnings.
641    
642    2.  Removed a bashism from the RunTest script.
643    
644    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
645        to unsigned type, result still unsigned" that was given by an MS compiler
646        on encountering the code "-sizeof(xxx)".
647    
648    4.  Partial matching support is added to the JIT compiler.
649    
650    5.  Fixed several bugs concerned with partial matching of items that consist
651        of more than one character:
652    
653        (a) /^(..)\1/ did not partially match "aba" because checking references was
654            done on an "all or nothing" basis. This also applied to repeated
655            references.
656    
657        (b) \R did not give a hard partial match if \r was found at the end of the
658            subject.
659    
660        (c) \X did not give a hard partial match after matching one or more
661            characters at the end of the subject.
662    
663        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
664            a partial match for the string "\r".
665    
666        (e) When newline was set to CRLF, the metacharacter "." did not recognize
667            a partial match for a CR character at the end of the subject string.
668    
669    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
670        running pcretest, the text "(JIT)" added to the output whenever JIT is
671        actually used to run the match.
672    
673    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
674        or /S+[+] with a digit between 1 and 7.
675    
676    8.  OP_NOT now supports any UTF character not just single-byte ones.
677    
678    9.  (*MARK) control verb is now supported by the JIT compiler.
679    
680    10. The command "./RunTest list" lists the available tests without actually
681        running any of them. (Because I keep forgetting what they all are.)
682    
683    11. Add PCRE_INFO_MAXLOOKBEHIND.
684    
685    12. Applied a (slightly modified) user-supplied patch that improves performance
686        when the heap is used for recursion (compiled with --disable-stack-for-
687        recursion). Instead of malloc and free for each heap frame each time a
688        logical recursion happens, frames are retained on a chain and re-used where
689        possible. This sometimes gives as much as 30% improvement.
690    
691    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
692        call.
693    
694    14. As documented, (*COMMIT) is now confined to within a positive assertion.
695    
696    15. It is now possible to link pcretest with libedit as an alternative to
697        libreadline.
698    
699    16. (*COMMIT) control verb is now supported by the JIT compiler.
700    
701    17. The Unicode data tables have been updated to Unicode 6.1.0.
702    
703    18. Added --file-list option to pcregrep.
704    
705    19. Added binary file support to pcregrep, including the -a, --binary-files,
706        -I, and --text options.
707    
708    20. The madvise function is renamed for posix_madvise for QNX compatibility
709        reasons. Fixed by Giuseppe D'Angelo.
710    
711    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
712    
713    22. Changed the option for creating symbolic links for 16-bit man pages from
714        -s to -sf so that re-installing does not cause issues.
715    
716    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
717    
718    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
719        matches in certain environments (the workspace was not being correctly
720        retained). Also added to pcre_dfa_exec() a simple plausibility check on
721        some of the workspace data at the beginning of a restart.
722    
723    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
724        was not doing so when it should - probably a typo introduced by SVN 528
725        (change 8.10/14).
726    
727    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
728        \w+ when the character tables indicated that \x{c4} was a word character.
729        There were several related cases, all because the tests for doing a table
730        lookup were testing for characters less than 127 instead of 255.
731    
732    27. If a pattern contains capturing parentheses that are not used in a match,
733        their slots in the ovector are set to -1. For those that are higher than
734        any matched groups, this happens at the end of processing. In the case when
735        there were back references that the ovector was too small to contain
736        (causing temporary malloc'd memory to be used during matching), and the
737        highest capturing number was not used, memory off the end of the ovector
738        was incorrectly being set to -1. (It was using the size of the temporary
739        memory instead of the true size.)
740    
741    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
742        ovector size, it uses memory at the end of the block that it has got.
743    
744    29. Check for an overlong MARK name and give an error at compile time. The
745        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
746    
747    30. JIT compiler update.
748    
749    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
750        Rill for the patch.
751    
752    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
753    
754    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
755    
756    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
757        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
758        was enabled.
759    
760    35. Improve JIT code generation for greedy plus quantifier.
761    
762    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
763        1 to "aa" instead of to an empty string. The bug affected repeated groups
764        that could potentially match an empty string.
765    
766    37. Optimizing single character iterators in JIT.
767    
768    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
769        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
770        codepoints that are too big for the mode are faulted, and in a UTF mode,
771        disallowed codepoints are also faulted.
772    
773    39. If PCRE was compiled with UTF support, in three places in the DFA
774        matcher there was code that should only have been obeyed in UTF mode, but
775        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
776        processing when bytes with values greater than 127 were present. In 16-bit
777        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
778        both cases the values are those that cannot be the first data item in a UTF
779        character. The three items that might have provoked this were recursions,
780        possessively repeated groups, and atomic groups.
781    
782    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
783        and pcregrep, because some OS require shared objects to be explicitly
784        passed to ld, causing the link step to fail if they are not.
785    
786    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
787        mode, patterns that started with \h* or \R* might be incorrectly matched.
788    
789    
790    Version 8.30 04-February-2012
791    -----------------------------
792    
793    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
794        name is defined in ctype.h.
795    
796    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
797        only in quite long subpatterns.
798    
799    3.  Removed the function pcre_info(), which has been obsolete and deprecated
800        since it was replaced by pcre_fullinfo() in February 2000.
801    
802    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
803        match a (*MARK), and the match failed at the start of the subject, a
804        reference to memory before the start of the subject could occur. This bug
805        was introduced by fix 17 of release 8.21.
806    
807    5.  A reference to an unset group with zero minimum repetition was giving
808        totally wrong answers (in non-JavaScript-compatibility mode). For example,
809        /(another)?(\1?)test/ matched against "hello world test". This bug was
810        introduced in release 8.13.
811    
812    6.  Add support for 16-bit character strings (a large amount of work involving
813        many changes and refactorings).
814    
815    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
816        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
817        from a file.
818    
819    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
820        rounding is not applied in this particular case).
821    
822    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
823        if they appear, or are escaped, in patterns.
824    
825    10. Get rid of a number of -Wunused-but-set-variable warnings.
826    
827    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
828        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
829        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
830        also returns the mark "x". This bug applied to capturing parentheses,
831        non-capturing parentheses, and atomic parentheses. It also applied to some
832        assertions.
833    
834    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
835        information out of configure.ac instead of relying on pcre.h.generic, which
836        is not stored in the repository.
837    
838    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
839        -lreadline.
840    
841    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
842    
843    15. Applied Graycode's patch to put the top-level frame on the stack rather
844        than the heap when not using the stack for recursion. This gives a
845        performance improvement in many cases when recursion is not deep.
846    
847    16. Experimental code added to "pcretest -C" to output the stack frame size.
848    
849    
850    Version 8.21 12-Dec-2011
851    ------------------------
852    
853    1.  Updating the JIT compiler.
854    
855    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
856        are added as well.
857    
858    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
859        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
860        calling _pcre_jit_exec. Some extra comments are added.
861    
862    4.  (*MARK) settings inside atomic groups that do not contain any capturing
863        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
864        was introduced by change 18 for 8.20.
865    
866    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
867        ECMA-262 standard.
868    
869    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
870        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
871        This bug was probably introduced by change 9 of 8.13.
872    
873    7.  While fixing 6 above, I noticed that a number of other items were being
874        incorrectly rejected as "not fixed length". This arose partly because newer
875        opcodes had not been added to the fixed-length checking code. I have (a)
876        corrected the bug and added tests for these items, and (b) arranged for an
877        error to occur if an unknown opcode is encountered while checking for fixed
878        length instead of just assuming "not fixed length". The items that were
879        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
880        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
881        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
882    
883    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
884        being incorrectly compiled and would have given unpredicatble results.
885    
886    9.  A possessively repeated subpattern with minimum repeat count greater than
887        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
888        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
889        the first (A) could occur when it should not.
890    
891    10. Add a cast and remove a redundant test from the code.
892    
893    11. JIT should use pcre_malloc/pcre_free for allocation.
894    
895    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
896        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
897        is anything other than /usr, -L is still shown).
898    
899    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
900    
901    14. Perl does not support \N without a following name in a [] class; PCRE now
902        also gives an error.
903    
904    15. If a forward reference was repeated with an upper limit of around 2000,
905        it caused the error "internal error: overran compiling workspace". The
906        maximum number of forward references (including repeats) was limited by the
907        internal workspace, and dependent on the LINK_SIZE. The code has been
908        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
909        the default depends on LINK_SIZE. There is a new upper limit (for safety)
910        of around 200,000 forward references. While doing this, I also speeded up
911        the filling in of repeated forward references.
912    
913    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
914        incorrectly expecting the subject to contain another "a" after the start.
915    
916    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
917        in the match, the SKIP should be ignored. This was not happening; instead
918        the SKIP was being treated as NOMATCH. For patterns such as
919        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
920        tested.
921    
922    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
923        now much more compatible with Perl, in particular in cases where the result
924        is a non-match for a non-anchored pattern. For example, if
925        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
926        "m", where previously it did not return a name. A side effect of this
927        change is that for partial matches, the last encountered mark name is
928        returned, as for non matches. A number of tests that were previously not
929        Perl-compatible have been moved into the Perl-compatible test files. The
930        refactoring has had the pleasing side effect of removing one argument from
931        the match() function, thus reducing its stack requirements.
932    
933    19. If the /S+ option was used in pcretest to study a pattern using JIT,
934        subsequent uses of /S (without +) incorrectly behaved like /S+.
935    
936    21. Retrieve executable code size support for the JIT compiler and fixing
937        some warnings.
938    
939    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
940        not work when the shorter character appeared right at the end of the
941        subject string.
942    
943    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
944        systems.
945    
946    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
947        output it when the /M option is used in pcretest.
948    
949    25. The CheckMan script was not being included in the distribution. Also, added
950        an explicit "perl" to run Perl scripts from the PrepareRelease script
951        because this is reportedly needed in Windows.
952    
953    26. If study data was being save in a file and studying had not found a set of
954        "starts with" bytes for the pattern, the data written to the file (though
955        never used) was taken from uninitialized memory and so caused valgrind to
956        complain.
957    
958    27. Updated RunTest.bat as provided by Sheri Pierce.
959    
960    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
961    
962    29. Computation of memory usage for the table of capturing group names was
963        giving an unnecessarily large value.
964    
965    
966    Version 8.20 21-Oct-2011
967    ------------------------
968    
969    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
970        a POSIX class. After further experiments with Perl, which convinced me that
971        Perl has bugs and confusions, a closing square bracket is no longer allowed
972        in a POSIX name. This bug also affected patterns with classes that started
973        with full stops.
974    
975    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
976        captured substring, but while checking the failing first alternative,
977        substring 1 is temporarily captured. If the output vector supplied to
978        pcre_exec() was not big enough for this capture, the yield of the function
979        was still zero ("insufficient space for captured substrings"). This cannot
980        be totally fixed without adding another stack variable, which seems a lot
981        of expense for a edge case. However, I have improved the situation in cases
982        such as /(a)(b)x|abc/ matched against "abc", where the return code
983        indicates that fewer than the maximum number of slots in the ovector have
984        been set.
985    
986    3.  Related to (2) above: when there are more back references in a pattern than
987        slots in the output vector, pcre_exec() uses temporary memory during
988        matching, and copies in the captures as far as possible afterwards. It was
989        using the entire output vector, but this conflicts with the specification
990        that only 2/3 is used for passing back captured substrings. Now it uses
991        only the first 2/3, for compatibility. This is, of course, another edge
992        case.
993    
994    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
995        main code base, and can be used by building with --enable-jit. When this is
996        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
997        runtime --no-jit option is given.
998    
999    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1000        ovector, the return from the function was zero, implying that there were
1001        other matches that did not fit. The correct "exactly full" value is now
1002        returned.
1003    
1004    6.  If a subpattern that was called recursively or as a subroutine contained
1005        (*PRUNE) or any other control that caused it to give a non-standard return,
1006        invalid errors such as "Error -26 (nested recursion at the same subject
1007        position)" or even infinite loops could occur.
1008    
1009    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1010        computing the minimum length on reaching *ACCEPT, and so ended up with the
1011        wrong value of 1 rather than 0. Further investigation indicates that
1012        computing a minimum subject length in the presence of *ACCEPT is difficult
1013        (think back references, subroutine calls), and so I have changed the code
1014        so that no minimum is registered for a pattern that contains *ACCEPT.
1015    
1016    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1017        it was not handled as intended. [But see 16 below.]
1018    
1019    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1020        Sheri Pierce.
1021    
1022    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1023        the first byte in a match must be "a".
1024    
1025    11. Change 17 for 8.13 increased the recursion depth for patterns like
1026        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1027        pattern contains any instances of (*THEN). If it does not, the old
1028        optimizations are restored. It would be nice to do this on a per-group
1029        basis, but at the moment that is not feasible.
1030    
1031    12. In some environments, the output of pcretest -C is CRLF terminated. This
1032        broke RunTest's code that checks for the link size. A single white space
1033        character after the value is now allowed for.
1034    
1035    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1036        For "fr", it uses the Windows-specific input and output files.
1037    
1038    14. If (*THEN) appeared in a group that was called recursively or as a
1039        subroutine, it did not work as intended. [But see next item.]
1040    
1041    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1042        pattern fragments (but not containing any | characters). If A and B are
1043        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1044        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1045        D. In other words, Perl considers parentheses that do not contain any |
1046        characters to be part of a surrounding alternative, whereas PCRE was
1047        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1048        differently. PCRE now behaves in the same way as Perl, except in the case
1049        of subroutine/recursion calls such as (?1) which have in any case always
1050        been different (but PCRE had them first :-).
1051    
1052    16. Related to 15 above: Perl does not treat the | in a conditional group as
1053        creating alternatives. Such a group is treated in the same way as an
1054        ordinary group without any | characters when processing (*THEN). PCRE has
1055        been changed to match Perl's behaviour.
1056    
1057    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1058        RunGrepTest script failed.
1059    
1060    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1061        inevitable for groups that contain captures, but it can lead to a lot of
1062        stack use in large patterns. The old behaviour has been restored for atomic
1063        groups that do not contain any capturing parentheses.
1064    
1065    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1066        suppress the check for a minimum subject length at run time. (If it was
1067        given to pcre_exec() or pcre_dfa_exec() it did work.)
1068    
1069    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1070        fail to work when decoding hex characters in data strings in EBCDIC
1071        environments.
1072    
1073    21. It appears that in at least one Mac OS environment, the isxdigit() function
1074        is implemented as a macro that evaluates to its argument more than once,
1075        contravening the C 90 Standard (I haven't checked a later standard). There
1076        was an instance in pcretest which caused it to go wrong when processing
1077        \x{...} escapes in subject strings. The has been rewritten to avoid using
1078        things like p++ in the argument of isxdigit().
1079    
1080    
1081    Version 8.13 16-Aug-2011
1082    ------------------------
1083    
1084    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1085    
1086    2.  Two minor typos in pcre_internal.h have been fixed.
1087    
1088    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1089        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1090        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1091    
1092    4.  There were a number of related bugs in the code for matching backrefences
1093        caselessly in UTF-8 mode when codes for the characters concerned were
1094        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1095        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1096        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1097        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1098        2-byte code at the end of the subject (it thought there wasn't enough data
1099        left).
1100    
1101    5.  Comprehensive information about what went wrong is now returned by
1102        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1103        as the output vector has at least 2 elements. The offset of the start of
1104        the failing character and a reason code are placed in the vector.
1105    
1106    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1107        now returned is for the first byte of the failing character, instead of the
1108        last byte inspected. This is an incompatible change, but I hope it is small
1109        enough not to be a problem. It makes the returned offset consistent with
1110        pcre_exec() and pcre_dfa_exec().
1111    
1112    7.  pcretest now gives a text phrase as well as the error number when
1113        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1114        failure, the offset and reason code are output.
1115    
1116    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1117        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1118        back over a single character (\n). This seems wrong (because it treated the
1119        two characters as a single entity when going forwards), conflicts with the
1120        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1121        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1122        has been changed.
1123    
1124    9.  Some internal refactoring has changed the processing so that the handling
1125        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1126        time (the PCRE_DOTALL option was changed this way some time ago: version
1127        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1128        which was always a bit of a fudge. It also means that there is one less
1129        argument for the match() function, which reduces its stack requirements
1130        slightly. This change also fixes an incompatibility with Perl: the pattern
1131        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1132    
1133    10. More internal refactoring has drastically reduced the number of recursive
1134        calls to match() for possessively repeated groups such as (abc)++ when
1135        using pcre_exec().
1136    
1137    11. While implementing 10, a number of bugs in the handling of groups were
1138        discovered and fixed:
1139    
1140        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1141        (a|)*(?1) gave a compile-time internal error.
1142        ((a|)+)+  did not notice that the outer group could match an empty string.
1143        (^a|^)+   was not marked as anchored.
1144        (.*a|.*)+ was not marked as matching at start or after a newline.
1145    
1146    12. Yet more internal refactoring has removed another argument from the match()
1147        function. Special calls to this function are now indicated by setting a
1148        value in a variable in the "match data" data block.
1149    
1150    13. Be more explicit in pcre_study() instead of relying on "default" for
1151        opcodes that mean there is no starting character; this means that when new
1152        ones are added and accidentally left out of pcre_study(), testing should
1153        pick them up.
1154    
1155    14. The -s option of pcretest has been documented for ages as being an old
1156        synonym of -m (show memory usage). I have changed it to mean "force study
1157        for every regex", that is, assume /S for every regex. This is similar to -i
1158        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1159        using it. It makes it easier to run collections of tests with and without
1160        study enabled, and thereby test pcre_study() more easily. All the standard
1161        tests are now run with and without -s (but some patterns can be marked as
1162        "never study" - see 20 below).
1163    
1164    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1165        restoration of the capturing data to the outer values was not happening
1166        correctly.
1167    
1168    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1169        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1170        pattern had matched an empty string, and so incorrectly returned a no
1171        match.
1172    
1173    17. There was optimizing code for the last branch of non-capturing parentheses,
1174        and also for the obeyed branch of a conditional subexpression, which used
1175        tail recursion to cut down on stack usage. Unfortunately, now that there is
1176        the possibility of (*THEN) occurring in these branches, tail recursion is
1177        no longer possible because the return has to be checked for (*THEN). These
1178        two optimizations have therefore been removed. [But see 8.20/11 above.]
1179    
1180    18. If a pattern containing \R was studied, it was assumed that \R always
1181        matched two bytes, thus causing the minimum subject length to be
1182        incorrectly computed because \R can also match just one byte.
1183    
1184    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1185        was incorrectly computed.
1186    
1187    20. If /S is present twice on a test pattern in pcretest input, it now
1188        *disables* studying, thereby overriding the use of -s on the command line
1189        (see 14 above). This is necessary for one or two tests to keep the output
1190        identical in both cases.
1191    
1192    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1193        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1194    
1195    22. When an atomic group that contained a capturing parenthesis was
1196        successfully matched, but the branch in which it appeared failed, the
1197        capturing was not being forgotten if a higher numbered group was later
1198        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1199        group 1 to "a", when in fact it should be unset. This applied to multi-
1200        branched capturing and non-capturing groups, repeated or not, and also to
1201        positive assertions (capturing in negative assertions does not happen
1202        in PCRE) and also to nested atomic groups.
1203    
1204    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1205        subject after a captured substring, to make it easier to tell which of a
1206        number of identical substrings has been captured.
1207    
1208    24. The way atomic groups are processed by pcre_exec() has been changed so that
1209        if they are repeated, backtracking one repetition now resets captured
1210        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1211        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1212        "aaa". Previously, it would have been "a". As part of this code
1213        refactoring, the way recursive calls are handled has also been changed.
1214    
1215    25. If an assertion condition captured any substrings, they were not passed
1216        back unless some other capturing happened later. For example, if
1217        (?(?=(a))a) was matched against "a", no capturing was returned.
1218    
1219    26. When studying a pattern that contained subroutine calls or assertions,
1220        the code for finding the minimum length of a possible match was handling
1221        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1222        group 1 called group 2 while simultaneously a separate group 2 called group
1223        1). A stack overflow occurred in this case. I have fixed this by limiting
1224        the recursion depth to 10.
1225    
1226    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1227        Fortmann. This supports explicit test numbers on the command line, and has
1228        argument validation and error reporting.
1229    
1230    28. An instance of \X with an unlimited repeat could fail if at any point the
1231        first character it looked at was a mark character.
1232    
1233    29. Some minor code refactoring concerning Unicode properties and scripts
1234        should reduce the stack requirement of match() slightly.
1235    
1236    30. Added the '=' option to pcretest to check the setting of unused capturing
1237        slots at the end of the pattern, which are documented as being -1, but are
1238        not included in the return count.
1239    
1240    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1241        compiled something random. Now it gives a compile-time error (as does
1242        Perl).
1243    
1244    32. A *MARK encountered during the processing of a positive assertion is now
1245        recorded and passed back (compatible with Perl).
1246    
1247    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1248        had alternative anchored branches, the search for a second match in a line
1249        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1250        matched the line "0102" twice. The same bug affected patterns that started
1251        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1252        twice.
1253    
1254    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1255        does, and because of capturing effects, quantifying parenthesized
1256        assertions may at times be useful. Quantifiers are now allowed for
1257        parenthesized assertions.
1258    
1259    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1260    
1261    36. \g was being checked for fancy things in a character class, when it should
1262        just be a literal "g".
1263    
1264    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1265        appearance of a nested POSIX class supersedes an apparent external class.
1266        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1267        unescaped square brackets may also appear as part of class names. For
1268        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1269        more like Perl. (But see 8.20/1 above.)
1270    
1271    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1272        was because it thought it was \N{name}, which is not supported).
1273    
1274    39. Add minix to OS list not supporting the -S option in pcretest.
1275    
1276    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1277        cannot analyze patterns in sufficient detail to catch mutual recursions
1278        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1279        subgroup is called recursively as a subpattern for a second time at the
1280        same position in the subject string. In previous releases this might have
1281        been caught by the recursion limit, or it might have run out of stack.
1282    
1283    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1284        happen only once. PCRE was, however incorrectly giving a compile time error
1285        "recursive call could loop indefinitely" because it cannot analyze the
1286        pattern in sufficient detail. The compile time test no longer happens when
1287        PCRE is compiling a conditional subpattern, but actual runaway loops are
1288        now caught at runtime (see 40 above).
1289    
1290    42. It seems that Perl allows any characters other than a closing parenthesis
1291        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1292        has been changed to be the same.
1293    
1294    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1295        as not to get warnings when autogen.sh is called. Also changed
1296        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1297    
1298    44. To help people who use pcregrep to scan files containing exceedingly long
1299        lines, the following changes have been made:
1300    
1301        (a) The default value of the buffer size parameter has been increased from
1302            8K to 20K. (The actual buffer used is three times this size.)
1303    
1304        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1305            PCRE is built.
1306    
1307        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1308            to be set at run time.
1309    
1310        (d) Numerical values in pcregrep options can be followed by K or M, for
1311            example --buffer-size=50K.
1312    
1313        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1314            given and the return code is set to 2.
1315    
1316    45. Add a pointer to the latest mark to the callout data block.
1317    
1318    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1319        partial match of an empty string instead of no match. This was specific to
1320        the use of ".".
1321    
1322    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1323        complete match instead of a partial match. This bug was dependent on both
1324        the PCRE_UTF8 and PCRE_DOTALL options being set.
1325    
1326    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1327        starting byte set, because \b was not being ignored.
1328    
1329    
1330    Version 8.12 15-Jan-2011
1331    ------------------------
1332    
1333    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1334        checks for such things as part of the documentation building process.
1335    
1336    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1337        --match-limit and --recursion-limit options (added for 8.11). In
1338        particular, this made one of the standard tests fail. (The integer value
1339        went into the wrong half of a long int.)
1340    
1341    3.  If the --colour option was given to pcregrep with -v (invert match), it
1342        did strange things, either producing crazy output, or crashing. It should,
1343        of course, ignore a request for colour when reporting lines that do not
1344        match.
1345    
1346    4.  Another pcregrep bug caused similar problems if --colour was specified with
1347        -M (multiline) and the pattern match finished with a line ending.
1348    
1349    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1350        matched in multiline mode, the following line was shown as part of the
1351        match. This seems wrong, so I have changed it.
1352    
1353    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1354        the check for further matches in the same line (so they could be coloured)
1355        to overrun the end of the current line. If another match was found, it was
1356        incorrectly shown (and then shown again when found in the next line).
1357    
1358    7.  If pcregrep was compiled under Windows, there was a reference to the
1359        function pcregrep_exit() before it was defined. I am assuming this was
1360        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1361        reported by a user. I've moved the definition above the reference.
1362    
1363    
1364    Version 8.11 10-Dec-2010
1365    ------------------------
1366    
1367    1.  (*THEN) was not working properly if there were untried alternatives prior
1368        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1369        backtracked to try for "b" instead of moving to the next alternative branch
1370        at the same level (in this case, to look for "c"). The Perl documentation
1371        is clear that when (*THEN) is backtracked onto, it goes to the "next
1372        alternative in the innermost enclosing group".
1373    
1374    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1375        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1376        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1377        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1378        (*THEN).
1379    
1380    3.  If \s appeared in a character class, it removed the VT character from
1381        the class, even if it had been included by some previous item, for example
1382        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1383        of \s, but is part of the POSIX "space" class.)
1384    
1385    4.  A partial match never returns an empty string (because you can always
1386        match an empty string at the end of the subject); however the checking for
1387        an empty string was starting at the "start of match" point. This has been
1388        changed to the "earliest inspected character" point, because the returned
1389        data for a partial match starts at this character. This means that, for
1390        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1391        (previously it gave "no match").
1392    
1393    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1394        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1395        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1396        has an implication that the given string is incomplete (because a partial
1397        match is preferred over a full match). For this reason, these items now
1398        give a partial match in this situation. [Aside: previously, the one case
1399        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1400        match rather than a full match, which was wrong by the old rules, but is
1401        now correct.]
1402    
1403    6.  There was a bug in the handling of #-introduced comments, recognized when
1404        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1405        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1406        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1407        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1408        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1409        places in pcre_compile().
1410    
1411    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1412        comments when looking ahead for named forward references to subpatterns,
1413        the only newline sequence it recognized was NL. It now handles newlines
1414        according to the set newline convention.
1415    
1416    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1417        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1418        cater for a lack of strerror(). These oversights have been fixed.
1419    
1420    9.  Added --match-limit and --recursion-limit to pcregrep.
1421    
1422    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1423    
1424    11. When the -o option was used, pcregrep was setting a return code of 1, even
1425        when matches were found, and --line-buffered was not being honoured.
1426    
1427    12. Added an optional parentheses number to the -o and --only-matching options
1428        of pcregrep.
1429    
1430    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1431        can match an empty string. The code to do it in pcretest and pcredemo
1432        needed fixing:
1433    
1434        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1435            only one byte after an empty string match just before CRLF (this case
1436            just got forgotten; "any" and "anycrlf" were OK).
1437    
1438        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1439            mode when an empty string match preceded an ASCII character followed by
1440            a non-ASCII character. (The code for advancing by one character rather
1441            than one byte was nonsense.)
1442    
1443        (c) The pcredemo.c sample program did not have any code at all to handle
1444            the cases when CRLF is a valid newline sequence.
1445    
1446    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1447        as a starting offset was within the subject string. There is now a new
1448        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1449        negative or greater than the length of the string. In order to test this,
1450        pcretest is extended to allow the setting of negative starting offsets.
1451    
1452    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1453        starting offset points to the beginning of a UTF-8 character was
1454        unnecessarily clumsy. I tidied it up.
1455    
1456    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1457        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1458    
1459    17. Nobody had reported that the --include_dir option, which was added in
1460        release 7.7 should have been called --include-dir (hyphen, not underscore)
1461        for compatibility with GNU grep. I have changed it to --include-dir, but
1462        left --include_dir as an undocumented synonym, and the same for
1463        --exclude-dir, though that is not available in GNU grep, at least as of
1464        release 2.5.4.
1465    
1466    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1467        characters from a string of bytes) have been redefined so as not to use
1468        loops, in order to improve performance in some environments. At the same
1469        time, I abstracted some of the common code into auxiliary macros to save
1470        repetition (this should not affect the compiled code).
1471    
1472    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1473        compile-time error is now given if \c is not followed by an ASCII
1474        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1475        different, and any byte value is allowed.)
1476    
1477    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1478        START_OPTIMIZE option, which is now allowed at compile time - but just
1479        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1480        to pcregrep and other applications that have no direct access to PCRE
1481        options. The new /Y option in pcretest sets this option when calling
1482        pcre_compile().
1483    
1484    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1485        back references. Groups containing recursive back references were forced to
1486        be atomic by that change, but in the case of named groups, the amount of
1487        memory required was incorrectly computed, leading to "Failed: internal
1488        error: code overflow". This has been fixed.
1489    
1490    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1491        pcretest.c, to avoid build problems in some Borland environments.
1492    
1493    
1494    Version 8.10 25-Jun-2010
1495    ------------------------
1496    
1497    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1498        THEN.
1499    
1500    2.  (*ACCEPT) was not working when inside an atomic group.
1501    
1502    3.  Inside a character class, \B is treated as a literal by default, but
1503        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1504        causes the error). The code is unchanged, but I tidied the documentation.
1505    
1506    4.  Inside a character class, PCRE always treated \R and \X as literals,
1507        whereas Perl faults them if its -w option is set. I have changed PCRE so
1508        that it faults them when PCRE_EXTRA is set.
1509    
1510    5.  Added support for \N, which always matches any character other than
1511        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1512    
1513    6.  When compiling pcregrep with newer versions of gcc which may have
1514        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1515        declared with attribute warn_unused_result" were given. Just casting the
1516        result to (void) does not stop the warnings; a more elaborate fudge is
1517        needed. I've used a macro to implement this.
1518    
1519    7.  Minor change to pcretest.c to avoid a compiler warning.
1520    
1521    8.  Added four artifical Unicode properties to help with an option to make
1522        \s etc use properties (see next item). The new properties are: Xan
1523        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1524    
1525    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1526        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1527        this option. Modified pcretest to add /W to test this facility. Added
1528        REG_UCP to make it available via the POSIX interface.
1529    
1530    10. Added --line-buffered to pcregrep.
1531    
1532    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1533        studied, and the match started with a letter with a code point greater than
1534        127 whose first byte was different to the first byte of the other case of
1535        the letter, the other case of this starting letter was not recognized
1536        (#976).
1537    
1538    12. If a pattern that was studied started with a repeated Unicode property
1539        test, for example, \p{Nd}+, there was the theoretical possibility of
1540        setting up an incorrect bitmap of starting bytes, but fortunately it could
1541        not have actually happened in practice until change 8 above was made (it
1542        added property types that matched character-matching opcodes).
1543    
1544    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1545        possible starting bytes for non-anchored patterns.
1546    
1547    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1548        \R, and also a number of cases that involve Unicode properties, both
1549        explicit and implicit when PCRE_UCP is set.
1550    
1551    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1552        input, it could crash or give wrong results if characters with values
1553        greater than 0xc0 were present in the subject string. (Detail: it assumed
1554        UTF-8 input when processing these items.)
1555    
1556    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1557        size_t is 64-bit (#991).
1558    
1559    17. Added a check for running out of memory when PCRE is compiled with
1560        --disable-stack-for-recursion (#990).
1561    
1562    18. If the last data line in a file for pcretest does not have a newline on
1563        the end, a newline was missing in the output.
1564    
1565    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1566        less than 128) in its various bitmaps. However, there is a facility for
1567        generating tables according to the current locale when PCRE is compiled. It
1568        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1569        characters, are recognized by isspace() and therefore were getting set in
1570        these tables, and indeed these tables seem to approximate to ISO 8859. This
1571        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1572        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1573        which of course cannot start UTF-8 characters. I have changed the code so
1574        that only real ASCII characters (less than 128) and the correct starting
1575        bytes for UTF-8 encodings are set for characters greater than 127 when in
1576        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1577        altogether.)
1578    
1579    20. Added the /T option to pcretest so as to be able to run tests with non-
1580        standard character tables, thus making it possible to include the tests
1581        used for 19 above in the standard set of tests.
1582    
1583    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1584        reference to a subpattern the other side of a comment that contains an
1585        opening parenthesis caused either an internal compiling error, or a
1586        reference to the wrong subpattern.
1587    
1588    
1589    Version 8.02 19-Mar-2010
1590    ------------------------
1591    
1592    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1593    
1594    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1595        configured.
1596    
1597    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1598        original author of that file, following a query about its status.
1599    
1600    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1601        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1602    
1603    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1604        quantifier applied to a forward-referencing subroutine call, could compile
1605        incorrect code or give the error "internal error: previously-checked
1606        referenced subpattern not found".
1607    
1608    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1609        variables to point to external functions. For these systems, therefore,
1610        pcre_malloc etc. are now initialized to local functions that call the
1611        relevant global functions.
1612    
1613    7.  There were two entries missing in the vectors called coptable and poptable
1614        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1615        I've fixed the data, and added a kludgy way of testing at compile time that
1616        the lengths are correct (equal to the number of opcodes).
1617    
1618    8.  Following on from 7, I added a similar kludge to check the length of the
1619        eint vector in pcreposix.c.
1620    
1621    9.  Error texts for pcre_compile() are held as one long string to avoid too
1622        much relocation at load time. To find a text, the string is searched,
1623        counting zeros. There was no check for running off the end of the string,
1624        which could happen if a new error number was added without updating the
1625        string.
1626    
1627    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1628    
1629    11. \K was not working if it appeared in an atomic group or in a group that
1630        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1631        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1632        the assertion is positive, but not if it is negative.
1633    
1634    12. Change 11 fortuitously reduced the size of the stack frame used in the
1635        "match()" function of pcre_exec.c by one pointer. Forthcoming
1636        implementation of support for (*MARK) will need an extra pointer on the
1637        stack; I have reserved it now, so that the stack frame size does not
1638        decrease.
1639    
1640    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1641        item in branch that calls a recursion is a subroutine call - as in the
1642        second branch in the above example - was incorrectly given the compile-
1643        time error "recursive call could loop indefinitely" because pcre_compile()
1644        was not correctly checking the subroutine for matching a non-empty string.
1645    
1646    14. The checks for overrunning compiling workspace could trigger after an
1647        overrun had occurred. This is a "should never occur" error, but it can be
1648        triggered by pathological patterns such as hundreds of nested parentheses.
1649        The checks now trigger 100 bytes before the end of the workspace.
1650    
1651    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1652    
1653    
1654    Version 8.01 19-Jan-2010
1655    ------------------------
1656    
1657    1.  If a pattern contained a conditional subpattern with only one branch (in
1658        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1659        computed the wrong minimum data length (which is of course zero for such
1660        subpatterns). This could cause incorrect "no match" results.
1661    
1662    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1663        the pattern is reset in the first branch, pcre_compile() failed with
1664        "internal error: code overflow at offset...". This happened only when
1665        the reset was to the original external option setting. (An optimization
1666        abstracts leading options settings into an external setting, which was the
1667        cause of this.)
1668    
1669    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1670        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1671        assertion pattern did not match (meaning that the assertion was true), it
1672        was incorrectly treated as false if the SKIP had been reached during the
1673        matching. This also applied to assertions used as conditions.
1674    
1675    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1676        assertion subpattern, including such a pattern used as a condition,
1677        unpredictable results occurred, instead of the error return
1678        PCRE_ERROR_DFA_UITEM.
1679    
1680    5.  The C++ GlobalReplace function was not working like Perl for the special
1681        situation when an empty string is matched. It now does the fancy magic
1682        stuff that is necessary.
1683    
1684    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1685        removed. (These were left over from very, very early versions of PCRE.)
1686    
1687    7.  Some cosmetic changes to the code to make life easier when compiling it
1688        as part of something else:
1689    
1690        (a) Change DEBUG to PCRE_DEBUG.
1691    
1692        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1693            called "current" as "current_branch", to prevent a collision with the
1694            Linux macro when compiled as a kernel module.
1695    
1696        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1697            prevent a collision with the Linux macro when compiled as a kernel
1698            module.
1699    
1700    8.  In pcre_compile() there are some checks for integer overflows that used to
1701        cast potentially large values to (double). This has been changed to that
1702        when building, a check for int64_t is made, and if it is found, it is used
1703        instead, thus avoiding the use of floating point arithmetic. (There is no
1704        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1705        double.
1706    
1707    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1708        2005 (difference between two addresses compared to an unsigned value).
1709    
1710    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1711        custom one, because of the following reported problem in Windows:
1712    
1713          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1714              under Win32.
1715          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1716              therefore missing the function definition.
1717          - The compiler thus generates a "C" signature for the test function.
1718          - The linker fails to find the "C" function.
1719          - PCRE fails to configure if asked to do so against libbz2.
1720    
1721    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1722        messages were output:
1723    
1724          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1725          rerunning libtoolize, to keep the correct libtool macros in-tree.
1726          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1727    
1728        I have done both of these things.
1729    
1730    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1731        most of the time, it *can* run out if it is given a pattern that contains a
1732        runaway infinite recursion. I updated the discussion in the pcrestack man
1733        page.
1734    
1735    13. Now that we have gone to the x.xx style of version numbers, the minor
1736        version may start with zero. Using 08 or 09 is a bad idea because users
1737        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1738        interpreted as invalid octal numbers. I've updated the previous comment in
1739        configure.ac, and also added a check that gives an error if 08 or 09 are
1740        used.
1741    
1742    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1743        causing partial matching to fail when the end of the subject matched \W
1744        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1745    
1746    15. There were some discrepancies between the declarations in pcre_internal.h
1747        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1748        their definitions. The declarations used "const uschar *" and the
1749        definitions used USPTR. Even though USPTR is normally defined as "const
1750        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1751        reported that: "This difference in casting confuses some C++ compilers, for
1752        example, SunCC recognizes above declarations as different functions and
1753        generates broken code for hbpcre." I have changed the declarations to use
1754        USPTR.
1755    
1756    16. GNU libtool is named differently on some systems. The autogen.sh script now
1757        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1758        (FreeBSD).
1759    
1760    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1761        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1762        comment: "Figure out how to create a longlong from a string: strtoll and
1763        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1764        instance, but it only takes 2 args instead of 3!"
1765    
1766    18. A subtle bug concerned with back references has been fixed by a change of
1767        specification, with a corresponding code fix. A pattern such as
1768        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1769        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1770        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1771        same bug. Such groups have to be quantified to be useful, or contained
1772        inside another quantified group. (If there's no repetition, the reference
1773        can never match.) The problem arises because, having left the group and
1774        moved on to the rest of the pattern, a later failure that backtracks into
1775        the group uses the captured value from the final iteration of the group
1776        rather than the correct earlier one. I have fixed this in PCRE by forcing
1777        any group that contains a reference to itself to be an atomic group; that
1778        is, there cannot be any backtracking into it once it has completed. This is
1779        similar to recursive and subroutine calls.
1780    
1781    
1782    Version 8.00 19-Oct-09
1783    ----------------------
1784    
1785    1.  The table for translating pcre_compile() error codes into POSIX error codes
1786        was out-of-date, and there was no check on the pcre_compile() error code
1787        being within the table. This could lead to an OK return being given in
1788        error.
1789    
1790    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1791        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1792        in a Windows environment.
1793    
1794    3.  The pcregrep --count option prints the count for each file even when it is
1795        zero, as does GNU grep. However, pcregrep was also printing all files when
1796        --files-with-matches was added. Now, when both options are given, it prints
1797        counts only for those files that have at least one match. (GNU grep just
1798        prints the file name in this circumstance, but including the count seems
1799        more useful - otherwise, why use --count?) Also ensured that the
1800        combination -clh just lists non-zero counts, with no names.
1801    
1802    4.  The long form of the pcregrep -F option was incorrectly implemented as
1803        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1804        but it seems right to fix it, and I didn't think it was worth preserving
1805        the old behaviour.
1806    
1807    5.  The command line items --regex=pattern and --regexp=pattern were not
1808        recognized by pcregrep, which required --regex pattern or --regexp pattern
1809        (with a space rather than an '='). The man page documented the '=' forms,
1810        which are compatible with GNU grep; these now work.
1811    
1812    6.  No libpcreposix.pc file was created for pkg-config; there was just
1813        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1814    
1815    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1816        when UCP support is not needed, by modifying the Python script that
1817        generates it from Unicode data files. This should not matter if the module
1818        is correctly used as a library, but I received one complaint about 50K of
1819        unwanted data. My guess is that the person linked everything into his
1820        program rather than using a library. Anyway, it does no harm.
1821    
1822    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1823        was a minimum greater than 1 for a wide character in a possessive
1824        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1825        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1826        character. Chaos in the form of incorrect output or a compiling loop could
1827        result.
1828    
1829    9.  The restrictions on what a pattern can contain when partial matching is
1830        requested for pcre_exec() have been removed. All patterns can now be
1831        partially matched by this function. In addition, if there are at least two
1832        slots in the offset vector, the offset of the earliest inspected character
1833        for the match and the offset of the end of the subject are set in them when
1834        PCRE_ERROR_PARTIAL is returned.
1835    
1836    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1837        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1838        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1839        and may be more useful for multi-segment matching.
1840    
1841    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1842        used to be given if ever the end of the subject was reached; now it is
1843        given only if matching could not proceed because another character was
1844        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1845        string "Z", which now yields "no match" instead of "partial match". In the
1846        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1847        final character ended with (*FAIL).
1848    
1849    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1850        if the pattern had a "must contain" character that was already found in the
1851        earlier partial match, unless partial matching was again requested. For
1852        example, with the pattern /dog.(body)?/, the "must contain" character is
1853        "g". If the first part-match was for the string "dog", restarting with
1854        "sbody" failed. This bug has been fixed.
1855    
1856    13. The string returned by pcre_dfa_exec() after a partial match has been
1857        changed so that it starts at the first inspected character rather than the
1858        first character of the match. This makes a difference only if the pattern
1859        starts with a lookbehind assertion or \b or \B (\K is not supported by
1860        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1861        matching functions compatible, and I think it's the right thing to do.
1862    
1863    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1864        so that the demonstration program is easily available in environments where
1865        PCRE has not been installed from source.
1866    
1867    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1868        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1869        library.
1870    
1871    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1872        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1873        is not the first non-POSIX option to be added. Clearly some people find
1874        these options useful.
1875    
1876    17. If a caller to the POSIX matching function regexec() passes a non-zero
1877        value for nmatch with a NULL value for pmatch, the value of
1878        nmatch is forced to zero.
1879    
1880    18. RunGrepTest did not have a test for the availability of the -u option of
1881        the diff command, as RunTest does. It now checks in the same way as
1882        RunTest, and also checks for the -b option.
1883    
1884    19. If an odd number of negated classes containing just a single character
1885        interposed, within parentheses, between a forward reference to a named
1886        subpattern and the definition of the subpattern, compilation crashed with
1887        an internal error, complaining that it could not find the referenced
1888        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1889        [The bug was that it was starting one character too far in when skipping
1890        over the character class, thus treating the ] as data rather than
1891        terminating the class. This meant it could skip too much.]
1892    
1893    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1894        /g option in pcretest when the pattern contains \K, which makes it possible
1895        to have an empty string match not at the start, even when the pattern is
1896        anchored. Updated pcretest and pcredemo to use this option.
1897    
1898    21. If the maximum number of capturing subpatterns in a recursion was greater
1899        than the maximum at the outer level, the higher number was returned, but
1900        with unset values at the outer level. The correct (outer level) value is
1901        now given.
1902    
1903    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1904        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1905        make it do so. The string so far is captured, making this feature
1906        compatible with Perl.
1907    
1908    23. The tests have been re-organized, adding tests 11 and 12, to make it
1909        possible to check the Perl 5.10 features against Perl 5.10.
1910    
1911    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1912        pattern matches a fixed length string. PCRE did not allow this; now it
1913        does. Neither allows recursion.
1914    
1915    25. I finally figured out how to implement a request to provide the minimum
1916        length of subject string that was needed in order to match a given pattern.
1917        (It was back references and recursion that I had previously got hung up
1918        on.) This code has now been added to pcre_study(); it finds a lower bound
1919        to the length of subject needed. It is not necessarily the greatest lower
1920        bound, but using it to avoid searching strings that are too short does give
1921        some useful speed-ups. The value is available to calling programs via
1922        pcre_fullinfo().
1923    
1924    26. While implementing 25, I discovered to my embarrassment that pcretest had
1925        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1926        study optimizations had never been tested with that matching function.
1927        Oops. What is worse, even when it was passed study data, there was a bug in
1928        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1929        were also very few tests of studied patterns with pcre_dfa_exec().
1930    
1931    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1932        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1933        on the other side of the coin, they are no longer allowed to have different
1934        names, because these cannot be distinguished in PCRE, and this has caused
1935        confusion. (This is a difference from Perl.)
1936    
1937    28. When duplicate subpattern names are present (necessarily with different
1938        numbers, as required by 27 above), and a test is made by name in a
1939        conditional pattern, either for a subpattern having been matched, or for
1940        recursion in such a pattern, all the associated numbered subpatterns are
1941        tested, and the overall condition is true if the condition is true for any
1942        one of them. This is the way Perl works, and is also more like the way
1943        testing by number works.
1944    
1945    
1946    Version 7.9 11-Apr-09
1947    ---------------------
1948    
1949    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1950        (pcretest), all targets were linked against these libraries. This included
1951        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1952        libraries. This caused unwanted dependencies to be created. This problem
1953        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1954        pcretest is linked with readline.
1955    
1956    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1957        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1958        moved outside it again, because FALSE and TRUE are already defined in AIX,
1959        but BOOL is not.
1960    
1961    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1962        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1963    
1964    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1965        hyphens) following file names and line numbers when outputting matching
1966        lines. This is not true; no spaces are inserted. I have also clarified the
1967        wording for the --colour (or --color) option.
1968    
1969    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1970        was not coloured; this is different to GNU grep, so I have changed it to be
1971        the same.
1972    
1973    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1974        each matching line was coloured. Now it goes on to look for further matches
1975        of any of the test patterns, which is the same behaviour as GNU grep.
1976    
1977    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1978        doesn't make sense to accept an empty string match in pcregrep, so I have
1979        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1980        seems to be how GNU grep behaves. [But see later change 40 for release
1981        8.33.]
1982    
1983    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1984        start or after a newline", because the conditional assertion was not being
1985        correctly handled. The rule now is that both the assertion and what follows
1986        in the first alternative must satisfy the test.
1987    
1988    9.  If auto-callout was enabled in a pattern with a conditional group whose
1989        condition was an assertion, PCRE could crash during matching, both with
1990        pcre_exec() and pcre_dfa_exec().
1991    
1992    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1993        used for matching.
1994    
1995    11. Unicode property support in character classes was not working for
1996        characters (bytes) greater than 127 when not in UTF-8 mode.
1997    
1998    12. Added the -M command line option to pcretest.
1999    
2000    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2001    
2002    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2003    
2004    16. Added comments and documentation about mis-use of no_arg in the C++
2005        wrapper.
2006    
2007    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2008        from Martin Jerabek that uses macro names for all relevant character and
2009        string constants.
2010    
2011    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2012        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2013        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2014        these, but not everybody uses configure.
2015    
2016    19. A conditional group that had only one branch was not being correctly
2017        recognized as an item that could match an empty string. This meant that an
2018        enclosing group might also not be so recognized, causing infinite looping
2019        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2020        with the subject "ab", where knowledge that the repeated group can match
2021        nothing is needed in order to break the loop.
2022    
2023    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2024        exec(), but without supplying a callout function, matching went wrong.
2025    
2026    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2027        leak if the size of the offset vector was greater than 30. When the vector
2028        is smaller, the saved offsets during recursion go onto a local stack
2029        vector, but for larger vectors malloc() is used. It was failing to free
2030        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2031        error, in fact).
2032    
2033    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2034        heapframe that is used only when UTF-8 support is enabled. This caused no
2035        problem, but was untidy.
2036    
2037    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2038        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2039        included within another project.
2040    
2041    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2042        slightly modified by me:
2043    
2044          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2045              not building pcregrep.
2046    
2047          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2048              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2049    
2050    25. Forward references, both numeric and by name, in patterns that made use of
2051        duplicate group numbers, could behave incorrectly or give incorrect errors,
2052        because when scanning forward to find the reference group, PCRE was not
2053        taking into account the duplicate group numbers. A pattern such as
2054        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2055    
2056    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2057        the feature of a custom pointer more persuasive (as requested by a user).
2058    
2059    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2060        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2061        pre-defined.
2062    
2063    28. Added support for (*UTF8) at the start of a pattern.
2064    
2065    29. Arrange for flags added by the "release type" setting in CMake to be shown
2066        in the configuration summary.
2067    
2068    
2069    Version 7.8 05-Sep-08
2070    ---------------------
2071    
2072    1.  Replaced UCP searching code with optimized version as implemented for Ad
2073        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2074        stage table and inline lookup instead of a function, giving speed ups of 2
2075        to 5 times on some simple patterns that I tested. Permission was given to
2076        distribute the MultiStage2.py script that generates the tables (it's not in
2077        the tarball, but is in the Subversion repository).
2078    
2079    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2080        scripts.
2081    
2082    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2083        a group with a zero qualifier. The result of the study could be incorrect,
2084        or the function might crash, depending on the pattern.
2085    
2086    4.  Caseless matching was not working for non-ASCII characters in back
2087        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2088        It now works when Unicode Property Support is available.
2089    
2090    5.  In pcretest, an escape such as \x{de} in the data was always generating
2091        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2092        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2093        truncation.
2094    
2095    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2096    
2097    7.  Added two (int) casts to pcregrep when printing the difference of two
2098        pointers, in case they are 64-bit values.
2099    
2100    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2101        test 2 if it fails.
2102    
2103    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2104        and a #define of that name to empty if it is not externally set. This is to
2105        allow users of MSVC to set it if necessary.
2106    
2107    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2108        the convenience functions in the pcre_get.c source file.
2109    
2110    11. An option change at the start of a pattern that had top-level alternatives
2111        could cause overwriting and/or a crash. This command provoked a crash in
2112        some environments:
2113    
2114          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2115    
2116        This potential security problem was recorded as CVE-2008-2371.
2117    
2118    12. For a pattern where the match had to start at the beginning or immediately
2119        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2120        pcre_dfa_exec() could read past the end of the passed subject if there was
2121        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2122        pcretest so that it places the subject at the end of its malloc-ed buffer.
2123    
2124    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2125        exec() might read past the end of the data buffer in UTF-8 mode.
2126    
2127    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2128        the data contained the byte 0x85 as part of a UTF-8 character within its
2129        first line. This applied both to normal and DFA matching.
2130    
2131    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2132        /^[^d]*?$/8 failed to match "abc".
2133    
2134    16. Added a missing copyright notice to pcrecpp_internal.h.
2135    
2136    17. Make it more clear in the documentation that values returned from
2137        pcre_exec() in ovector are byte offsets, not character counts.
2138    
2139    18. Tidied a few places to stop certain compilers from issuing warnings.
2140    
2141    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2142        supplied by Stefan Weber. I made a further small update for 7.8 because
2143        there is a change of source arrangements: the pcre_searchfuncs.c module is
2144        replaced by pcre_ucd.c.
2145    
2146    
2147    Version 7.7 07-May-08
2148    ---------------------
2149    
2150    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2151        a string to a long long, pretend we don't even have a long long." This is
2152        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2153    
2154    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2155        pre-7.6 versions, which defined a global no_arg variable instead of putting
2156        it in the RE class. (See also #8 below.)
2157    
2158    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2159        Lopes.
2160    
2161    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2162    
2163        (1) The include/exclude patterns were being applied to the whole pathnames
2164            of files, instead of just to the final components.
2165    
2166        (2) If there was more than one level of directory, the subdirectories were
2167            skipped unless they satisfied the include/exclude conditions. This is
2168            inconsistent with GNU grep (and could even be seen as contrary to the
2169            pcregrep specification - which I improved to make it absolutely clear).
2170            The action now is always to scan all levels of directory, and just
2171            apply the include/exclude patterns to regular files.
2172    
2173    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2174        --exclude_dir in the tests to avoid scanning .svn directories.
2175    
2176    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2177        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2178        doesn't support NULs in patterns.
2179    
2180    7.  Added some missing "const"s to declarations of static tables in
2181        pcre_compile.c and pcre_dfa_exec.c.
2182    
2183    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2184        caused by fix #2  above. (Subsequently also a second patch to fix the
2185        first patch. And a third patch - this was a messy problem.)
2186    
2187    9.  Applied Craig's patch to remove the use of push_back().
2188    
2189    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2190        matching function regexec().
2191    
2192    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2193        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2194        references. PCRE supports relative numbers with this syntax (I don't think
2195        Oniguruma does).
2196    
2197    12. Previously, a group with a zero repeat such as (...){0} was completely
2198        omitted from the compiled regex. However, this means that if the group
2199        was called as a subroutine from elsewhere in the pattern, things went wrong
2200        (an internal error was given). Such groups are now left in the compiled
2201        pattern, with a new opcode that causes them to be skipped at execution
2202        time.
2203    
2204    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2205        to the way PCRE behaves:
2206    
2207        (a) A lone ] character is dis-allowed (Perl treats it as data).
2208    
2209        (b) A back reference to an unmatched subpattern matches an empty string
2210            (Perl fails the current match path).
2211    
2212        (c) A data ] in a character class must be notated as \] because if the
2213            first data character in a class is ], it defines an empty class. (In
2214            Perl it is not possible to have an empty class.) The empty class []
2215            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2216            The negative empty class [^] matches any one character, independently
2217            of the DOTALL setting.
2218    
2219    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2220        non-existent subpattern following a character class starting with ']' and
2221        containing () gave an internal compiling error instead of "reference to
2222        non-existent subpattern". Fortunately, when the pattern did exist, the
2223        compiled code was correct. (When scanning forwards to check for the
2224        existence of the subpattern, it was treating the data ']' as terminating
2225        the class, so got the count wrong. When actually compiling, the reference
2226        was subsequently set up correctly.)
2227    
2228    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2229        it was being rejected as not supported by pcre_dfa_exec(), even though
2230        other assertions are supported. I have made pcre_dfa_exec() support
2231        (*FAIL).
2232    
2233    16. The implementation of 13c above involved the invention of a new opcode,
2234        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2235        cannot be changed at match time, I realized I could make a small
2236        improvement to matching performance by compiling OP_ALLANY instead of
2237        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2238        on the OP_ANY path.
2239    
2240    17. Compiling pcretest on Windows with readline support failed without the
2241        following two fixes: (1) Make the unistd.h include conditional on
2242        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2243    
2244    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2245        ncurses library to be included for pcretest when ReadLine support is
2246        requested, but also to allow for it to be overridden. This patch came from
2247        Daniel Bergström.
2248    
2249    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2250        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2251        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2252        spotting this.
2253    
2254    
2255    Version 7.6 28-Jan-08
2256    ---------------------
2257    
2258    1.  A character class containing a very large number of characters with
2259        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2260        overflow.
2261    
2262    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2263        HAVE_LONG_LONG is not defined.
2264    
2265    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2266        bring them up to date and include new features. This patch includes:
2267    
2268        - Fixed PH's badly added libz and libbz2 support.
2269        - Fixed a problem with static linking.
2270        - Added pcredemo. [But later removed - see 7 below.]
2271        - Fixed dftables problem and added an option.
2272        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2273            HAVE_LONG_LONG.
2274        - Added readline support for pcretest.
2275        - Added an listing of the option settings after cmake has run.
2276    
2277    4.  A user submitted a patch to Makefile that makes it easy to create
2278        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2279        Makefile.am that cause it to include this special target, without
2280        affecting anything else. Note that the same mingw target plus all
2281        the other distribution libraries and programs are now supported
2282        when configuring with CMake (see 6 below) instead of with
2283        Configure/Make.
2284    
2285    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2286        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2287        exported in the Windows port". It has not yet been confirmed that the patch
2288        solves the problem, but it does no harm.
2289    
2290    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2291        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2292        with CMake, and also correct the comment about stack recursion.
2293    
2294    7.  Remove the automatic building of pcredemo from the ./configure system and
2295        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2296        of a program that users should build themselves after PCRE is installed, so
2297        building it automatically is not really right. What is more, it gave
2298        trouble in some build environments.
2299    
2300    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2301    
2302    
2303    Version 7.5 10-Jan-08
2304    ---------------------
2305    
2306    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2307        values in parens when parsing an RE using the C++ wrapper."
2308    
2309    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2310        Characters greater than 255 were excluded from the class instead of being
2311        included.
2312    
2313    3.  The same bug as (2) above applied to negated POSIX classes such as
2314        [:^space:].
2315    
2316    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2317        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2318        I have changed it.
2319    
2320    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2321        first named subpattern) and a construct such as (?&a) would reference the
2322        first named subpattern whose name started with "a" (in other words, the
2323        length check was missing). Both these problems are fixed. "Subpattern name
2324        expected" is now given for (?&) (a zero-length name), and this patch also
2325        makes it give the same error for \k'' (previously it complained that that
2326        was a reference to a non-existent subpattern).
2327    
2328    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2329        this is right because (?- can be followed by option settings as well as by
2330        digits. I have, however, made the messages clearer.
2331    
2332    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2333        than the number used in the conditional) now cause a compile-time error.
2334        This is actually not compatible with Perl, which accepts such patterns, but
2335        treats the conditional as always being FALSE (as PCRE used to), but it
2336        seems to me that giving a diagnostic is better.
2337    
2338    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2339        and messages.
2340    
2341    9.  Fix two occurrences of "backslash" in comments that should have been
2342        "backspace".
2343    
2344    10. Remove two redundant lines of code that can never be obeyed (their function
2345        was moved elsewhere).
2346    
2347    11. The program that makes PCRE's Unicode character property table had a bug
2348        which caused it to generate incorrect table entries for sequences of
2349        characters that have the same character type, but are in different scripts.
2350        It amalgamated them into a single range, with the script of the first of
2351        them. In other words, some characters were in the wrong script. There were
2352        thirteen such cases, affecting characters in the following ranges:
2353    
2354          U+002b0 - U+002c1
2355          U+0060c - U+0060d
2356          U+0061e - U+00612
2357          U+0064b - U+0065e
2358          U+0074d - U+0076d
2359          U+01800 - U+01805
2360          U+01d00 - U+01d77
2361          U+01d9b - U+01dbf
2362          U+0200b - U+0200f
2363          U+030fc - U+030fe
2364          U+03260 - U+0327f
2365          U+0fb46 - U+0fbb1
2366          U+10450 - U+1049d
2367    
2368    12. The -o option (show only the matching part of a line) for pcregrep was not
2369        compatible with GNU grep in that, if there was more than one match in a
2370        line, it showed only the first of them. It now behaves in the same way as
2371        GNU grep.
2372    
2373    13. If the -o and -v options were combined for pcregrep, it printed a blank
2374        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2375        does the same. The return code can be used to tell if there were any
2376        non-matching lines.
2377    
2378    14. Added --file-offsets and --line-offsets to pcregrep.
2379    
2380    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2381        infinitely looping recursion. The bug was that positive lookaheads were not
2382        being skipped when checking for a possible empty match (negative lookaheads
2383        and both kinds of lookbehind were skipped).
2384    
2385    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2386        inclusion of <windows.h> to before rather than after the definition of
2387        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2388    
2389    17. Specifying a possessive quantifier with a specific limit for a Unicode
2390        character property caused pcre_compile() to compile bad code, which led at
2391        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2392        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2393        caused the error; without that there was no problem.
2394    
2395    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2396    
2397    19. Added --enable-pcretest-libreadline.
2398    
2399    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2400        RE::GlobalReplace(). As a result, the number of replacements returned was
2401        double what it should be. I removed one of the increments, but Craig sent a
2402        later patch that removed the other one (the right fix) and added unit tests
2403        that check the return values (which was not done before).
2404    
2405    21. Several CMake things:
2406    
2407        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2408            the names libpcre and libpcreposix, not just pcre and pcreposix.
2409    
2410        (2) The above change means that pcretest and pcregrep are now correctly
2411            linked with the newly-built libraries, not previously installed ones.
2412    
2413        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2414    
2415    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2416        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2417        UTF-8 newline character). The key issue is that the pattern starts .*;
2418        this means that the match must be either at the beginning, or after a
2419        newline. The bug was in the code for advancing after a failed match and
2420        checking that the new position followed a newline. It was not taking
2421        account of UTF-8 characters correctly.
2422    
2423    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2424        character classes. PCRE was not treating the sequence [:...:] as a
2425        character class unless the ... were all letters. Perl, however, seems to
2426        allow any characters between [: and :], though of course it rejects as
2427        unknown any "names" that contain non-letters, because all the known class
2428        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2429        for example, whereas PCRE did not - it did not recognize a POSIX character
2430        class. This seemed a bit dangerous, so the code has been changed to be
2431        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2432        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2433        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2434        Perl does, and where it didn't before.
2435    
2436    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2437        Windows environments %n is disabled by default.
2438    
2439    
2440    Version 7.4 21-Sep-07
2441    ---------------------
2442    
2443    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2444        means that a class such as [\s] counted as "explicit reference to CR or
2445        LF". That isn't really right - the whole point of the change was to try to
2446        help when there was an actual mention of one of the two characters. So now
2447        the change happens only if \r or \n (or a literal CR or LF) character is
2448        encountered.
2449    
2450    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2451        of both had grown to the point where there were only 3 bits left.
2452        Fortunately, there was spare space in the data structure, and so I have
2453        moved the internal flags into a new 16-bit field to free up more option
2454        bits.
2455    
2456    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2457        but did not set the internal JCHANGED flag - either of these is enough to
2458        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2459        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2460        start sets both bits.
2461    
2462    4.  Added options (at build time, compile time, exec time) to change \R from
2463        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2464    
2465    5.  doc/pcresyntax.html was missing from the distribution.
2466    
2467    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2468        compatibility, even though it is no longer used.
2469    
2470    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2471        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2472        windows.h file is present (where different names are used). [This was
2473        reversed later after testing - see 16 below.]
2474    
2475    8.  Changed all #include <config.h> to #include "config.h". There were also
2476        some further <pcre.h> cases that I changed to "pcre.h".
2477    
2478    9.  When pcregrep was used with the --colour option, it missed the line ending
2479        sequence off the lines that it output.
2480    
2481    10. It was pointed out to me that arrays of string pointers cause lots of
2482        relocations when a shared library is dynamically loaded. A technique of
2483        using a single long string with a table of offsets can drastically reduce
2484        these. I have refactored PCRE in four places to do this. The result is
2485        dramatic:
2486    
2487          Originally:                          290
2488          After changing UCP table:            187
2489          After changing error message table:   43
2490          After changing table of "verbs"       36
2491          After changing table of Posix names   22
2492    
2493        Thanks to the folks working on Gregex for glib for this insight.
2494    
2495    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2496        unicode-properties was also set.
2497    
2498    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2499    
2500    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2501        checked only for CRLF.
2502    
2503    14. Added casts to pcretest.c to avoid compiler warnings.
2504    
2505    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2506    
2507    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2508        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2509        entirely. This removes changes made in 7 above.
2510    
2511    17. The CMake files have been updated, and there is now more information about
2512        building with CMake in the NON-UNIX-USE document.
2513    
2514    
2515    Version 7.3 28-Aug-07
2516    ---------------------
2517    
2518     1. In the rejigging of the build system that eventually resulted in 7.1, the
2519        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2520        brackets there is not right, since it causes compilers to look for an
2521        installed pcre.h, not the version that is in the source that is being
2522        compiled (which of course may be different). I have changed it back to:
2523    
2524          #include "pcre.h"
2525    
2526        I have a vague recollection that the change was concerned with compiling in
2527        different directories, but in the new build system, that is taken care of
2528        by the VPATH setting the Makefile.
2529    
2530     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2531        when the subject happened to end in the byte 0x85 (e.g. if the last
2532        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2533        characters but of course it shouldn't be taken as a newline when it is part
2534        of another character. The bug was that, for an unlimited repeat of . in
2535        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2536        characters when looking for a newline.
2537    
2538     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2539    
2540     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2541        in debug output.
2542    
2543     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2544        long printing in the pcrecpp unittest when running under MinGW.
2545    
2546     6. ESC_K was left out of the EBCDIC table.
2547    
2548     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2549        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2550        limit also applies to "virtual nesting" when a pattern is recursive, and in
2551        this case 1000 isn't so big. I have been able to remove this limit at the
2552        expense of backing off one optimization in certain circumstances. Normally,
2553        when pcre_exec() would call its internal match() function recursively and
2554        immediately return the result unconditionally, it uses a "tail recursion"
2555        feature to save stack. However, when a subpattern that can match an empty
2556        string has an unlimited repetition quantifier, it no longer makes this
2557        optimization. That gives it a stack frame in which to save the data for
2558        checking that an empty string has been matched. Previously this was taken
2559        from the 1000-entry workspace that had been reserved. So now there is no
2560        explicit limit, but more stack is used.
2561    
2562     8. Applied Daniel's patches to solve problems with the import/export magic
2563        syntax that is required for Windows, and which was going wrong for the
2564        pcreposix and pcrecpp parts of the library. These were overlooked when this
2565        problem was solved for the main library.
2566    
2567     9. There were some crude static tests to avoid integer overflow when computing
2568        the size of patterns that contain repeated groups with explicit upper
2569        limits. As the maximum quantifier is 65535, the maximum group length was
2570        set at 30,000 so that the product of these two numbers did not overflow a
2571        32-bit integer. However, it turns out that people want to use groups that
2572        are longer than 30,000 bytes (though not repeat them that many times).
2573        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2574        made it possible to implement the integer overflow checks in a much more
2575        dynamic way, which I have now done. The artificial limitation on group
2576        length has been removed - we now have only the limit on the total length of
2577        the compiled pattern, which depends on the LINK_SIZE setting.
2578    
2579    10. Fixed a bug in the documentation for get/copy named substring when
2580        duplicate names are permitted. If none of the named substrings are set, the
2581        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2582        empty string.
2583    
2584    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2585        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2586        because the ] is interpreted as the first data character and the
2587        terminating ] is not found. PCRE has been made compatible with Perl in this
2588        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2589        cause memory overwriting.
2590    
2591    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2592        string has been matched (to stop an infinite loop). It was not recognizing
2593        a conditional subpattern that could match an empty string if that
2594        subpattern was within another subpattern. For example, it looped when
2595        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2596        condition was not nested. This bug has been fixed.
2597    
2598    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2599        past the start of the subject in the presence of bytes with the top bit
2600        set, for example "\x8aBCD".
2601    
2602    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2603        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2604    
2605    14. Optimized (?!) to (*FAIL).
2606    
2607    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2608        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2609        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2610        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2611        does: it's just the validity check that is more restrictive.
2612    
2613    16. Inserted checks for integer overflows during escape sequence (backslash)
2614        processing, and also fixed erroneous offset values for syntax errors during
2615        backslash processing.
2616    
2617    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2618        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2619    
2620    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2621        caused an overrun.
2622    
2623    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2624        something other than just ASCII characters) inside a group that had an
2625        unlimited repeat caused a loop at compile time (while checking to see
2626        whether the group could match an empty string).
2627    
2628    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2629        [\P{Any}] did so. (Error in the code for printing property names.)
2630    
2631    21. An orphan \E inside a character class could cause a crash.
2632    
2633    22. A repeated capturing bracket such as (A)? could cause a wild memory
2634        reference during compilation.
2635    
2636    23. There are several functions in pcre_compile() that scan along a compiled
2637        expression for various reasons (e.g. to see if it's fixed length for look
2638        behind). There were bugs in these functions when a repeated \p or \P was
2639        present in the pattern. These operators have additional parameters compared
2640        with \d, etc, and these were not being taken into account when moving along
2641        the compiled data. Specifically:
2642    
2643        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2644            length.
2645    
2646        (b) An item such as \pL+ within a repeated group could cause crashes or
2647            loops.
2648    
2649        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2650            "reference to non-existent subpattern" error.
2651    
2652        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2653    
2654    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2655        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2656    
2657    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2658    
2659    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2660        character were causing crashes (broken optimization).
2661    
2662    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2663        \p or \P) caused a compile-time loop.
2664    
2665    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2666        break. For example, the unstudied pattern [\r\n]A does not match the string
2667        "\r\nA" because change 7.0/46 below moves the current point on by two
2668        characters after failing to match at the start. However, the pattern \nA
2669        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2670        the same is true. There doesn't seem any very clean way out of this, but
2671        what I have chosen to do makes the common cases work: PCRE now takes note
2672        of whether there can be an explicit match for \r or \n anywhere in the
2673        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2674        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2675        pattern has explicit CR or LF references.
2676    
2677    29. Added (*CR) etc for changing newline setting at start of pattern.
2678    
2679    
2680    Version 7.2 19-Jun-07
2681    ---------------------
2682    
2683     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2684        which is apparently normally available under Windows.
2685    
2686     2. Re-jig the pcregrep tests with different newline settings in an attempt
2687        to make them independent of the local environment's newline setting.
2688    
2689     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2690    
2691     4. Some of the "internals" tests were previously cut out when the link size
2692        was not 2, because the output contained actual offsets. The recent new
2693        "Z" feature of pcretest means that these can be cut out, making the tests
2694        usable with all link sizes.
2695    
2696     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2697        stack recursion. This gives a massive performance boost under BSD, but just
2698        a small improvement under Linux. However, it saves one field in the frame
2699        in all cases.
2700    
2701     6. Added more features from the forthcoming Perl 5.10:
2702    
2703        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2704            recursion call. It refers to the nth most recently opened parentheses.
2705    
2706        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2707            to be opened parentheses.
2708    
2709        (c) Conditions that refer to capturing parentheses can be specified
2710            relatively, for example, (?(-2)... or (?(+3)...
2711    
2712        (d) \K resets the start of the current match so that everything before
2713            is not part of it.
2714    
2715        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2716    
2717        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2718            reference syntax.
2719    
2720        (g) (?| introduces a group in which the numbering of parentheses in each
2721            alternative starts with the same number.
2722    
2723        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2724    
2725     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2726        PCRE_INFO_JCHANGED.
2727    
2728     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2729        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2730        for detecting groups that can match an empty string.
2731    
2732     9. A pattern with a very large number of alternatives (more than several
2733        hundred) was running out of internal workspace during the pre-compile
2734        phase, where pcre_compile() figures out how much memory will be needed. A
2735        bit of new cunning has reduced the workspace needed for groups with
2736        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2737        workspace instead of running out of the 4096 that are available.
2738    
2739    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2740    
2741    11. Applied patch from Google to remove an optimization that didn't quite work.
2742        The report of the bug said:
2743    
2744          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2745          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2746          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2747    
2748    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2749        it matched the wrong number of bytes.
2750    
2751    
2752    Version 7.1 24-Apr-07
2753  ---------------------  ---------------------
2754    
2755   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one   1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
# Line 33  Version 7.1 12-Mar-07 Line 2781  Version 7.1 12-Mar-07
2781   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan   5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2782      Weber: (1) pcre_internal.h was missing some function renames; (2) updated      Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2783      makevp.bat for the current PCRE, using the additional files      makevp.bat for the current PCRE, using the additional files
2784      makevp-compile.txt, makevp-linklib.txt, and pcregexp.pas.      makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2785    
2786   6. A Windows user reported a minor discrepancy with test 2, which turned out   6. A Windows user reported a minor discrepancy with test 2, which turned out
2787      to be caused by a trailing space on an input line that had got lost in his      to be caused by a trailing space on an input line that had got lost in his
# Line 116  Version 7.1 12-Mar-07 Line 2864  Version 7.1 12-Mar-07
2864  14. If erroroffset was passed as NULL to pcre_compile, it provoked a  14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2865      segmentation fault instead of returning the appropriate error message.      segmentation fault instead of returning the appropriate error message.
2866    
2867    15. In multiline mode when the newline sequence was set to "any", the pattern
2868        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2869        This doesn't seem right; it now treats the CRLF combination as the line
2870        ending, and so does not match in that case. It's only a pattern such as ^$
2871        that would hit this one: something like ^ABC$ would have failed after \r
2872        and then tried again after \r\n.
2873    
2874    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2875        in an attempt to make files that differ only in their line terminators
2876        compare equal. This works on Linux.
2877    
2878    17. Under certain error circumstances pcregrep might try to free random memory
2879        as it exited. This is now fixed, thanks to valgrind.
2880    
2881    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2882        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2883        was because its rules for how to advance for /g after matching an empty
2884        string at the end of a line did not allow for this case. They now check for
2885        it specially.
2886    
2887    20. pcretest is supposed to handle patterns and data of any length, by
2888        extending its buffers when necessary. It was getting this wrong when the
2889        buffer for a data line had to be extended.
2890    
2891    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2892        CRLF as a newline sequence.
2893    
2894    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2895        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2896        I have nevertheless tidied it up.
2897    
2898    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2899    
2900    24. Added a man page for pcre-config.
2901    
2902    
2903  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06
2904  ---------------------  ---------------------

Legend:
Removed from v.127  
changed lines
  Added in v.1450

  ViewVC Help
Powered by ViewVC 1.1.5