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

Diff of /code/trunk/ChangeLog

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

revision 131 by ph10, Mon Mar 26 15:10:12 2007 UTC revision 1464 by zherczeg, Fri Mar 14 11:45:56 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 04-March-2014
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    19. Little endian PowerPC systems are supported now by the JIT compiler.
98    
99    20. The fast forward newline mechanism could enter to an infinite loop on
100        certain invalid UTF-8 input. Although we don't support these cases
101        this issue can be fixed by a performance optimization.
102    
103    21. Change 33 of 8.34 is not sufficient to ensure stack safety because it does
104        not take account if existing stack usage. There is now a new global
105        variable called pcre_stack_guard that can be set to point to an external
106        function to check stack availability. It is called at the start of
107        processing every parenthesized group.
108    
109    22. A typo in the code meant that in ungreedy mode the max/min qualifier
110        behaved like a min-possessive qualifier, and, for example, /a{1,3}b/U did
111        not match "ab".
112    
113    23. When UTF was disabled, the JIT program reported some incorrect compile
114        errors. These messages are silenced now.
115    
116    
117    Version 8.34 15-December-2013
118    -----------------------------
119    
120    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
121        executable memory. Patch inspired by Carsten Klein.
122    
123    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
124        this macro is never tested and has no effect, because the work to support
125        coverage involves only compiling and linking options and special targets in
126        the Makefile. The comment in config.h implied that defining the macro would
127        enable coverage support, which is totally false. There was also support for
128        setting this macro in the CMake files (my fault, I just copied it from
129        configure). SUPPORT_GCOV has now been removed.
130    
131    3.  Make a small performance improvement in strlen16() and strlen32() in
132        pcretest.
133    
134    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
135        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
136        these cases). There was also one in pcretest.c.
137    
138    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
139    
140    6.  In UTF mode, the code for checking whether a group could match an empty
141        string (which is used for indefinitely repeated groups to allow for
142        breaking an infinite loop) was broken when the group contained a repeated
143        negated single-character class with a character that occupied more than one
144        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
145        UTF-8 mode). The effect was undefined: the group might or might not be
146        deemed as matching an empty string, or the program might have crashed.
147    
148    7.  The code for checking whether a group could match an empty string was not
149        recognizing that \h, \H, \v, \V, and \R must match a character.
150    
151    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
152        an empty string. If it can, pcretest shows this in its information output.
153    
154    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
155        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
156        matched by pcre_exec() without using JIT:
157    
158        (a) If the rest of the pattern did not match after a maximal run of
159            grapheme clusters, the code for backing up to try with fewer of them
160            did not always back up over a full grapheme when characters that do not
161            have the modifier quality were involved, e.g. Hangul syllables.
162    
163        (b) If the match point in a subject started with modifier character, and
164            there was no match, the code could incorrectly back up beyond the match
165            point, and potentially beyond the first character in the subject,
166            leading to a segfault or an incorrect match result.
167    
168    10. A conditional group with an assertion condition could lead to PCRE
169        recording an incorrect first data item for a match if no other first data
170        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
171        first data item, and therefore matched "ca" after "c" instead of at the
172        start.
173    
174    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
175        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
176    
177    12. The source of pcregrep now includes z/OS-specific code so that it can be
178        compiled for z/OS as part of the special z/OS distribution.
179    
180    13. Added the -T and -TM options to pcretest.
181    
182    14. The code in pcre_compile.c for creating the table of named capturing groups
183        has been refactored. Instead of creating the table dynamically during the
184        actual compiling pass, the information is remembered during the pre-compile
185        pass (on the stack unless there are more than 20 named groups, in which
186        case malloc() is used) and the whole table is created before the actual
187        compile happens. This has simplified the code (it is now nearly 150 lines
188        shorter) and prepared the way for better handling of references to groups
189        with duplicate names.
190    
191    15. A back reference to a named subpattern when there is more than one of the
192        same name now checks them in the order in which they appear in the pattern.
193        The first one that is set is used for the reference. Previously only the
194        first one was inspected. This change makes PCRE more compatible with Perl.
195    
196    16. Unicode character properties were updated from Unicode 6.3.0.
197    
198    17. The compile-time code for auto-possessification has been refactored, based
199        on a patch by Zoltan Herczeg. It now happens after instead of during
200        compilation. The code is cleaner, and more cases are handled. The option
201        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
202        options in pcretest are provided to set it. It can also be set by
203        (*NO_AUTO_POSSESS) at the start of a pattern.
204    
205    18. The character VT has been added to the default ("C" locale) set of
206        characters that match \s and are generally treated as white space,
207        following this same change in Perl 5.18. There is now no difference between
208        "Perl space" and "POSIX space". Whether VT is treated as white space in
209        other locales depends on the locale.
210    
211    19. The code for checking named groups as conditions, either for being set or
212        for being recursed, has been refactored (this is related to 14 and 15
213        above). Processing unduplicated named groups should now be as fast at
214        numerical groups, and processing duplicated groups should be faster than
215        before.
216    
217    20. Two patches to the CMake build system, by Alexander Barkov:
218    
219          (1) Replace the "source" command by "." in CMakeLists.txt because
220              "source" is a bash-ism.
221    
222          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
223              without these the CMake build does not work on Solaris.
224    
225    21. Perl has changed its handling of \8 and \9. If there is no previously
226        encountered capturing group of those numbers, they are treated as the
227        literal characters 8 and 9 instead of a binary zero followed by the
228        literals. PCRE now does the same.
229    
230    22. Following Perl, added \o{} to specify codepoints in octal, making it
231        possible to specify values greater than 0777 and also making them
232        unambiguous.
233    
234    23. Perl now gives an error for missing closing braces after \x{... instead of
235        treating the string as literal. PCRE now does the same.
236    
237    24. RunTest used to grumble if an inappropriate test was selected explicitly,
238        but just skip it when running all tests. This make it awkward to run ranges
239        of tests when one of them was inappropriate. Now it just skips any
240        inappropriate tests, as it always did when running all tests.
241    
242    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
243        character types such as \d or \w, too many callouts were inserted, and the
244        data that they returned was rubbish.
245    
246    26. In UCP mode, \s was not matching two of the characters that Perl matches,
247        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
248        were matched by \h. The code has now been refactored so that the lists of
249        the horizontal and vertical whitespace characters used for \h and \v (which
250        are defined only in one place) are now also used for \s.
251    
252    27. Add JIT support for the 64 bit TileGX architecture.
253        Patch by Jiong Wang (Tilera Corporation).
254    
255    28. Possessive quantifiers for classes (both explicit and automatically
256        generated) now use special opcodes instead of wrapping in ONCE brackets.
257    
258    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
259        (because it's meaningless), this was not happening when PCRE_CASELESS was
260        set. Not wrong, but inefficient.
261    
262    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
263        properties for \w, \d, etc) is present in a test regex. Otherwise if the
264        test contains no characters greater than 255, Perl doesn't realise it
265        should be using Unicode semantics.
266    
267    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
268        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
269        does in Unicode mode.
270    
271    32. Added the "forbid" facility to pcretest so that putting tests into the
272        wrong test files can sometimes be quickly detected.
273    
274    33. There is now a limit (default 250) on the depth of nesting of parentheses.
275        This limit is imposed to control the amount of system stack used at compile
276        time. It can be changed at build time by --with-parens-nest-limit=xxx or
277        the equivalent in CMake.
278    
279    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
280        errors. Perl warns for these when in warning mode, but PCRE has no facility
281        for giving warnings.
282    
283    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
284        However, this was not working for (?!) because it is optimized to (*FAIL),
285        for which PCRE does not allow quantifiers. The optimization is now disabled
286        when a quantifier follows (?!). I can't see any use for this, but it makes
287        things uniform.
288    
289    36. Perl no longer allows group names to start with digits, so I have made this
290        change also in PCRE. It simplifies the code a bit.
291    
292    37. In extended mode, Perl ignores spaces before a + that indicates a
293        possessive quantifier. PCRE allowed a space before the quantifier, but not
294        before the possessive +. It now does.
295    
296    38. The use of \K (reset reported match start) within a repeated possessive
297        group such as (a\Kb)*+ was not working.
298    
299    40. Document that the same character tables must be used at compile time and
300        run time, and that the facility to pass tables to pcre_exec() and
301        pcre_dfa_exec() is for use only with saved/restored patterns.
302    
303    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
304        features for Builds with MSVC:
305    
306        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
307           builds). The .rc files can be used to set FileDescription and many other
308           attributes.
309    
310        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
311           This allows higher-level build scripts which want .pdb files to avoid
312           hard-coding the exact files needed."
313    
314    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
315        mean "start of word" and "end of word", respectively, as a transition aid.
316    
317    43. A minimizing repeat of a class containing codepoints greater than 255 in
318        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
319        compiled to use the heap for recursion.
320    
321    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
322        is configured.
323    
324    
325    Version 8.33 28-May-2013
326    ------------------------
327    
328    1.  Added 'U' to some constants that are compared to unsigned integers, to
329        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
330        variables that are added to signed variables, to ensure the result is
331        signed and can be negated.
332    
333    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
334        CMake config files.
335    
336    3.  Revise the creation of config.h.generic so that all boolean macros are
337        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
338        overriding via -D on the command line possible.
339    
340    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
341        to unsigned int is reported to make a quite noticeable speed difference in
342        a specific Windows environment. Testing on Linux did also appear to show
343        some benefit (and it is clearly not harmful). Also fixed the definition of
344        Xop which should be unsigned.
345    
346    5.  Related to (4), changing the definition of the intermediate variable cc
347        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
348        improvements.
349    
350    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
351        unnecessary spaces.
352    
353    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
354        and later.
355    
356    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
357    
358    9.  Optimizing fast_forward_start_bits in JIT.
359    
360    10. Adding support for callouts in JIT, and fixing some issues revealed
361        during this work. Namely:
362    
363        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
364    
365        (b) Minimum length was not checked before the matching is started.
366    
367    11. The value of capture_last that is passed to callouts was incorrect in some
368        cases when there was a capture on one path that was subsequently abandoned
369        after a backtrack. Also, the capture_last value is now reset after a
370        recursion, since all captures are also reset in this case.
371    
372    12. The interpreter no longer returns the "too many substrings" error in the
373        case when an overflowing capture is in a branch that is subsequently
374        abandoned after a backtrack.
375    
376    13. In the pathological case when an offset vector of size 2 is used, pcretest
377        now prints out the matched string after a yield of 0 or 1.
378    
379    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
380        Only supported by the JIT compiler at the moment.
381    
382    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
383    
384    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
385        offset of the starting point of the matching process, provided the offsets
386        vector is large enough.
387    
388    17. The \A escape now records a lookbehind value of 1, though its execution
389        does not actually inspect the previous character. This is to ensure that,
390        in partial multi-segment matching, at least one character from the old
391        segment is retained when a new segment is processed. Otherwise, if there
392        are no lookbehinds in the pattern, \A might match incorrectly at the start
393        of a new segment.
394    
395    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
396    
397    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
398        gives some modest performance improvement in 8-bit mode.
399    
400    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
401        be expressed in certain programming languages using Universal Character
402        Names.
403    
404    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
405        which points out that "non characters" are not "characters that may not
406        appear in Unicode strings" but rather "characters that are reserved for
407        internal use and have only local meaning".
408    
409    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
410        there was a conditional group that depended on an assertion, if the
411        assertion was false, the callout that immediately followed the alternation
412        in the condition was skipped when pcre_exec() was used for matching.
413    
414    23. Allow an explicit callout to be inserted before an assertion that is the
415        condition for a conditional group, for compatibility with automatic
416        callouts, which always insert a callout at this point.
417    
418    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
419        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
420    
421    25. (*PRUNE) is now supported by the JIT compiler.
422    
423    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
424    
425    27. Fix the case where there are two or more SKIPs with arguments that may be
426        ignored.
427    
428    28. (*SKIP) is now supported by the JIT compiler.
429    
430    29. (*THEN) is now supported by the JIT compiler.
431    
432    30. Update RunTest with additional test selector options.
433    
434    31. The way PCRE handles backtracking verbs has been changed in two ways.
435    
436        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
437        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
438        backtracking. In some cases this makes it more Perl-compatible, but Perl's
439        rather obscure rules do not always do the same thing.
440    
441        (2) Previously, backtracking verbs were confined within assertions. This is
442        no longer the case for positive assertions, except for (*ACCEPT). Again,
443        this sometimes improves Perl compatibility, and sometimes does not.
444    
445    32. A number of tests that were in test 2 because Perl did things differently
446        have been moved to test 1, because either Perl or PCRE has changed, and
447        these tests are now compatible.
448    
449    32. Backtracking control verbs are now handled in the same way in JIT and
450        interpreter.
451    
452    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
453        contained a forward subroutine reference caused a compile error.
454    
455    34. Auto-detect and optimize limited repetitions in JIT.
456    
457    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
458        blocking (*UTF) etc.
459    
460    36. In the interpreter, maximizing pattern repetitions for characters and
461        character types now use tail recursion, which reduces stack usage.
462    
463    37. The value of the max lookbehind was not correctly preserved if a compiled
464        and saved regex was reloaded on a host of different endianness.
465    
466    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
467        of the compiled pattern block, expand the flags field from 16 to 32 bits
468        because it was almost full.
469    
470    39. Try madvise first before posix_madvise.
471    
472    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
473        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
474        original change locked out all matches of empty strings. This has been
475        changed so that one match of an empty string per line is recognized.
476        Subsequent searches on the same line (for colouring or for --only-matching,
477        for example) do not recognize empty strings.
478    
479    41. Applied a user patch to fix a number of spelling mistakes in comments.
480    
481    42. Data lines longer than 65536 caused pcretest to crash.
482    
483    43. Clarified the data type for length and startoffset arguments for pcre_exec
484        and pcre_dfa_exec in the function-specific man pages, where they were
485        explicitly stated to be in bytes, never having been updated. I also added
486        some clarification to the pcreapi man page.
487    
488    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
489        a segmentation fault.
490    
491    
492    Version 8.32 30-November-2012
493    -----------------------------
494    
495    1.  Improved JIT compiler optimizations for first character search and single
496        character iterators.
497    
498    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
499        Patch by Daniel Richard G.
500    
501    3.  Single character iterator optimizations in the JIT compiler.
502    
503    4.  Improved JIT compiler optimizations for character ranges.
504    
505    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
506        Reported by Giuseppe D'Angelo.
507    
508    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
509        of a line, was being set incorrectly in cases where .* appeared inside
510        atomic brackets at the start of a pattern, or where there was a subsequent
511        *PRUNE or *SKIP.
512    
513    7.  Improved instruction cache flush for POWER/PowerPC.
514        Patch by Daniel Richard G.
515    
516    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
517        grep:
518    
519        (a) There is now no limit to the number of patterns to be matched.
520    
521        (b) An error is given if a pattern is too long.
522    
523        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
524            are now supported.
525    
526        (d) --exclude-from and --include-from (multiple use) have been added.
527    
528        (e) Exclusions and inclusions now apply to all files and directories, not
529            just to those obtained from scanning a directory recursively.
530    
531        (f) Multiple uses of -f and --file-list are now supported.
532    
533        (g) In a Windows environment, the default for -d has been changed from
534            "read" (the GNU grep default) to "skip", because otherwise the presence
535            of a directory in the file list provokes an error.
536    
537        (h) The documentation has been revised and clarified in places.
538    
539    9.  Improve the matching speed of capturing brackets.
540    
541    10. Changed the meaning of \X so that it now matches a Unicode extended
542        grapheme cluster.
543    
544    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
545        out POSIX threads when JIT support is configured.
546    
547    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
548    
549    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
550        structure could go wrong in environments where size_t is not the same size
551        as int.
552    
553    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
554        set.
555    
556    15. The EBCDIC support had decayed; later updates to the code had included
557        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
558        general tidy up of EBCDIC-related issues, and the documentation was also
559        not quite right. There is now a test that can be run on ASCII systems to
560        check some of the EBCDIC-related things (but is it not a full test).
561    
562    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
563        in a small tidy to the code.
564    
565    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
566    
567    18. If the --only-matching (-o) option in pcregrep is specified multiple
568        times, each one causes appropriate output. For example, -o1 -o2 outputs the
569        substrings matched by the 1st and 2nd capturing parentheses. A separating
570        string can be specified by --om-separator (default empty).
571    
572    19. Improving the first n character searches.
573    
574    20. Turn case lists for horizontal and vertical white space into macros so that
575        they are defined only once.
576    
577    21. This set of changes together give more compatible Unicode case-folding
578        behaviour for characters that have more than one other case when UCP
579        support is available.
580    
581        (a) The Unicode property table now has offsets into a new table of sets of
582            three or more characters that are case-equivalent. The MultiStage2.py
583            script that generates these tables (the pcre_ucd.c file) now scans
584            CaseFolding.txt instead of UnicodeData.txt for character case
585            information.
586    
587        (b) The code for adding characters or ranges of characters to a character
588            class has been abstracted into a generalized function that also handles
589            case-independence. In UTF-mode with UCP support, this uses the new data
590            to handle characters with more than one other case.
591    
592        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
593            whose other case is greater than 256 are now correctly matched
594            caselessly. Previously, the high codepoint matched the low one, but not
595            vice versa.
596    
597        (d) The processing of \h, \H, \v, and \ in character classes now makes use
598            of the new class addition function, using character lists defined as
599            macros alongside the case definitions of 20 above.
600    
601        (e) Caseless back references now work with characters that have more than
602            one other case.
603    
604        (f) General caseless matching of characters with more than one other case
605            is supported.
606    
607    22. Unicode character properties were updated from Unicode 6.2.0
608    
609    23. Improved CMake support under Windows. Patch by Daniel Richard G.
610    
611    24. Add support for 32-bit character strings, and UTF-32
612    
613    25. Major JIT compiler update (code refactoring and bugfixing).
614        Experimental Sparc 32 support is added.
615    
616    26. Applied a modified version of Daniel Richard G's patch to create
617        pcre.h.generic and config.h.generic by "make" instead of in the
618        PrepareRelease script.
619    
620    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
621        pcre_compile.c when checking for a zero character.
622    
623    28. Introducing a native interface for JIT. Through this interface, the compiled
624        machine code can be directly executed. The purpose of this interface is to
625        provide fast pattern matching, so several sanity checks are not performed.
626        However, feature tests are still performed. The new interface provides
627        1.4x speedup compared to the old one.
628    
629    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
630        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
631        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
632    
633    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
634        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
635        the "old" RFC 2279). Instead, it ended up passing a negative length to
636        pcre_exec().
637    
638    31. Add support for GCC's visibility feature to hide internal functions.
639    
640    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
641        "unknown -C option" after outputting 0 or 1.
642    
643    33. There is now support for generating a code coverage report for the test
644        suite in environments where gcc is the compiler and lcov is installed. This
645        is mainly for the benefit of the developers.
646    
647    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
648        unaddressable using valgrind annotations, allowing valgrind to detect
649        invalid memory accesses. This is mainly for the benefit of the developers.
650    
651    25. (*UTF) can now be used to start a pattern in any of the three libraries.
652    
653    26. Give configure error if --enable-cpp but no C++ compiler found.
654    
655    
656    Version 8.31 06-July-2012
657    -------------------------
658    
659    1.  Fixing a wrong JIT test case and some compiler warnings.
660    
661    2.  Removed a bashism from the RunTest script.
662    
663    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
664        to unsigned type, result still unsigned" that was given by an MS compiler
665        on encountering the code "-sizeof(xxx)".
666    
667    4.  Partial matching support is added to the JIT compiler.
668    
669    5.  Fixed several bugs concerned with partial matching of items that consist
670        of more than one character:
671    
672        (a) /^(..)\1/ did not partially match "aba" because checking references was
673            done on an "all or nothing" basis. This also applied to repeated
674            references.
675    
676        (b) \R did not give a hard partial match if \r was found at the end of the
677            subject.
678    
679        (c) \X did not give a hard partial match after matching one or more
680            characters at the end of the subject.
681    
682        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
683            a partial match for the string "\r".
684    
685        (e) When newline was set to CRLF, the metacharacter "." did not recognize
686            a partial match for a CR character at the end of the subject string.
687    
688    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
689        running pcretest, the text "(JIT)" added to the output whenever JIT is
690        actually used to run the match.
691    
692    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
693        or /S+[+] with a digit between 1 and 7.
694    
695    8.  OP_NOT now supports any UTF character not just single-byte ones.
696    
697    9.  (*MARK) control verb is now supported by the JIT compiler.
698    
699    10. The command "./RunTest list" lists the available tests without actually
700        running any of them. (Because I keep forgetting what they all are.)
701    
702    11. Add PCRE_INFO_MAXLOOKBEHIND.
703    
704    12. Applied a (slightly modified) user-supplied patch that improves performance
705        when the heap is used for recursion (compiled with --disable-stack-for-
706        recursion). Instead of malloc and free for each heap frame each time a
707        logical recursion happens, frames are retained on a chain and re-used where
708        possible. This sometimes gives as much as 30% improvement.
709    
710    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
711        call.
712    
713    14. As documented, (*COMMIT) is now confined to within a positive assertion.
714    
715    15. It is now possible to link pcretest with libedit as an alternative to
716        libreadline.
717    
718    16. (*COMMIT) control verb is now supported by the JIT compiler.
719    
720    17. The Unicode data tables have been updated to Unicode 6.1.0.
721    
722    18. Added --file-list option to pcregrep.
723    
724    19. Added binary file support to pcregrep, including the -a, --binary-files,
725        -I, and --text options.
726    
727    20. The madvise function is renamed for posix_madvise for QNX compatibility
728        reasons. Fixed by Giuseppe D'Angelo.
729    
730    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
731    
732    22. Changed the option for creating symbolic links for 16-bit man pages from
733        -s to -sf so that re-installing does not cause issues.
734    
735    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
736    
737    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
738        matches in certain environments (the workspace was not being correctly
739        retained). Also added to pcre_dfa_exec() a simple plausibility check on
740        some of the workspace data at the beginning of a restart.
741    
742    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
743        was not doing so when it should - probably a typo introduced by SVN 528
744        (change 8.10/14).
745    
746    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
747        \w+ when the character tables indicated that \x{c4} was a word character.
748        There were several related cases, all because the tests for doing a table
749        lookup were testing for characters less than 127 instead of 255.
750    
751    27. If a pattern contains capturing parentheses that are not used in a match,
752        their slots in the ovector are set to -1. For those that are higher than
753        any matched groups, this happens at the end of processing. In the case when
754        there were back references that the ovector was too small to contain
755        (causing temporary malloc'd memory to be used during matching), and the
756        highest capturing number was not used, memory off the end of the ovector
757        was incorrectly being set to -1. (It was using the size of the temporary
758        memory instead of the true size.)
759    
760    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
761        ovector size, it uses memory at the end of the block that it has got.
762    
763    29. Check for an overlong MARK name and give an error at compile time. The
764        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
765    
766    30. JIT compiler update.
767    
768    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
769        Rill for the patch.
770    
771    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
772    
773    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
774    
775    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
776        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
777        was enabled.
778    
779    35. Improve JIT code generation for greedy plus quantifier.
780    
781    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
782        1 to "aa" instead of to an empty string. The bug affected repeated groups
783        that could potentially match an empty string.
784    
785    37. Optimizing single character iterators in JIT.
786    
787    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
788        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
789        codepoints that are too big for the mode are faulted, and in a UTF mode,
790        disallowed codepoints are also faulted.
791    
792    39. If PCRE was compiled with UTF support, in three places in the DFA
793        matcher there was code that should only have been obeyed in UTF mode, but
794        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
795        processing when bytes with values greater than 127 were present. In 16-bit
796        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
797        both cases the values are those that cannot be the first data item in a UTF
798        character. The three items that might have provoked this were recursions,
799        possessively repeated groups, and atomic groups.
800    
801    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
802        and pcregrep, because some OS require shared objects to be explicitly
803        passed to ld, causing the link step to fail if they are not.
804    
805    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
806        mode, patterns that started with \h* or \R* might be incorrectly matched.
807    
808    
809    Version 8.30 04-February-2012
810    -----------------------------
811    
812    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
813        name is defined in ctype.h.
814    
815    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
816        only in quite long subpatterns.
817    
818    3.  Removed the function pcre_info(), which has been obsolete and deprecated
819        since it was replaced by pcre_fullinfo() in February 2000.
820    
821    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
822        match a (*MARK), and the match failed at the start of the subject, a
823        reference to memory before the start of the subject could occur. This bug
824        was introduced by fix 17 of release 8.21.
825    
826    5.  A reference to an unset group with zero minimum repetition was giving
827        totally wrong answers (in non-JavaScript-compatibility mode). For example,
828        /(another)?(\1?)test/ matched against "hello world test". This bug was
829        introduced in release 8.13.
830    
831    6.  Add support for 16-bit character strings (a large amount of work involving
832        many changes and refactorings).
833    
834    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
835        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
836        from a file.
837    
838    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
839        rounding is not applied in this particular case).
840    
841    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
842        if they appear, or are escaped, in patterns.
843    
844    10. Get rid of a number of -Wunused-but-set-variable warnings.
845    
846    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
847        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
848        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
849        also returns the mark "x". This bug applied to capturing parentheses,
850        non-capturing parentheses, and atomic parentheses. It also applied to some
851        assertions.
852    
853    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
854        information out of configure.ac instead of relying on pcre.h.generic, which
855        is not stored in the repository.
856    
857    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
858        -lreadline.
859    
860    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
861    
862    15. Applied Graycode's patch to put the top-level frame on the stack rather
863        than the heap when not using the stack for recursion. This gives a
864        performance improvement in many cases when recursion is not deep.
865    
866    16. Experimental code added to "pcretest -C" to output the stack frame size.
867    
868    
869    Version 8.21 12-Dec-2011
870    ------------------------
871    
872    1.  Updating the JIT compiler.
873    
874    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
875        are added as well.
876    
877    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
878        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
879        calling _pcre_jit_exec. Some extra comments are added.
880    
881    4.  (*MARK) settings inside atomic groups that do not contain any capturing
882        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
883        was introduced by change 18 for 8.20.
884    
885    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
886        ECMA-262 standard.
887    
888    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
889        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
890        This bug was probably introduced by change 9 of 8.13.
891    
892    7.  While fixing 6 above, I noticed that a number of other items were being
893        incorrectly rejected as "not fixed length". This arose partly because newer
894        opcodes had not been added to the fixed-length checking code. I have (a)
895        corrected the bug and added tests for these items, and (b) arranged for an
896        error to occur if an unknown opcode is encountered while checking for fixed
897        length instead of just assuming "not fixed length". The items that were
898        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
899        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
900        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
901    
902    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
903        being incorrectly compiled and would have given unpredicatble results.
904    
905    9.  A possessively repeated subpattern with minimum repeat count greater than
906        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
907        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
908        the first (A) could occur when it should not.
909    
910    10. Add a cast and remove a redundant test from the code.
911    
912    11. JIT should use pcre_malloc/pcre_free for allocation.
913    
914    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
915        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
916        is anything other than /usr, -L is still shown).
917    
918    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
919    
920    14. Perl does not support \N without a following name in a [] class; PCRE now
921        also gives an error.
922    
923    15. If a forward reference was repeated with an upper limit of around 2000,
924        it caused the error "internal error: overran compiling workspace". The
925        maximum number of forward references (including repeats) was limited by the
926        internal workspace, and dependent on the LINK_SIZE. The code has been
927        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
928        the default depends on LINK_SIZE. There is a new upper limit (for safety)
929        of around 200,000 forward references. While doing this, I also speeded up
930        the filling in of repeated forward references.
931    
932    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
933        incorrectly expecting the subject to contain another "a" after the start.
934    
935    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
936        in the match, the SKIP should be ignored. This was not happening; instead
937        the SKIP was being treated as NOMATCH. For patterns such as
938        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
939        tested.
940    
941    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
942        now much more compatible with Perl, in particular in cases where the result
943        is a non-match for a non-anchored pattern. For example, if
944        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
945        "m", where previously it did not return a name. A side effect of this
946        change is that for partial matches, the last encountered mark name is
947        returned, as for non matches. A number of tests that were previously not
948        Perl-compatible have been moved into the Perl-compatible test files. The
949        refactoring has had the pleasing side effect of removing one argument from
950        the match() function, thus reducing its stack requirements.
951    
952    19. If the /S+ option was used in pcretest to study a pattern using JIT,
953        subsequent uses of /S (without +) incorrectly behaved like /S+.
954    
955    21. Retrieve executable code size support for the JIT compiler and fixing
956        some warnings.
957    
958    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
959        not work when the shorter character appeared right at the end of the
960        subject string.
961    
962    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
963        systems.
964    
965    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
966        output it when the /M option is used in pcretest.
967    
968    25. The CheckMan script was not being included in the distribution. Also, added
969        an explicit "perl" to run Perl scripts from the PrepareRelease script
970        because this is reportedly needed in Windows.
971    
972    26. If study data was being save in a file and studying had not found a set of
973        "starts with" bytes for the pattern, the data written to the file (though
974        never used) was taken from uninitialized memory and so caused valgrind to
975        complain.
976    
977    27. Updated RunTest.bat as provided by Sheri Pierce.
978    
979    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
980    
981    29. Computation of memory usage for the table of capturing group names was
982        giving an unnecessarily large value.
983    
984    
985    Version 8.20 21-Oct-2011
986    ------------------------
987    
988    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
989        a POSIX class. After further experiments with Perl, which convinced me that
990        Perl has bugs and confusions, a closing square bracket is no longer allowed
991        in a POSIX name. This bug also affected patterns with classes that started
992        with full stops.
993    
994    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
995        captured substring, but while checking the failing first alternative,
996        substring 1 is temporarily captured. If the output vector supplied to
997        pcre_exec() was not big enough for this capture, the yield of the function
998        was still zero ("insufficient space for captured substrings"). This cannot
999        be totally fixed without adding another stack variable, which seems a lot
1000        of expense for a edge case. However, I have improved the situation in cases
1001        such as /(a)(b)x|abc/ matched against "abc", where the return code
1002        indicates that fewer than the maximum number of slots in the ovector have
1003        been set.
1004    
1005    3.  Related to (2) above: when there are more back references in a pattern than
1006        slots in the output vector, pcre_exec() uses temporary memory during
1007        matching, and copies in the captures as far as possible afterwards. It was
1008        using the entire output vector, but this conflicts with the specification
1009        that only 2/3 is used for passing back captured substrings. Now it uses
1010        only the first 2/3, for compatibility. This is, of course, another edge
1011        case.
1012    
1013    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
1014        main code base, and can be used by building with --enable-jit. When this is
1015        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
1016        runtime --no-jit option is given.
1017    
1018    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
1019        ovector, the return from the function was zero, implying that there were
1020        other matches that did not fit. The correct "exactly full" value is now
1021        returned.
1022    
1023    6.  If a subpattern that was called recursively or as a subroutine contained
1024        (*PRUNE) or any other control that caused it to give a non-standard return,
1025        invalid errors such as "Error -26 (nested recursion at the same subject
1026        position)" or even infinite loops could occur.
1027    
1028    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
1029        computing the minimum length on reaching *ACCEPT, and so ended up with the
1030        wrong value of 1 rather than 0. Further investigation indicates that
1031        computing a minimum subject length in the presence of *ACCEPT is difficult
1032        (think back references, subroutine calls), and so I have changed the code
1033        so that no minimum is registered for a pattern that contains *ACCEPT.
1034    
1035    8.  If (*THEN) was present in the first (true) branch of a conditional group,
1036        it was not handled as intended. [But see 16 below.]
1037    
1038    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
1039        Sheri Pierce.
1040    
1041    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
1042        the first byte in a match must be "a".
1043    
1044    11. Change 17 for 8.13 increased the recursion depth for patterns like
1045        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
1046        pattern contains any instances of (*THEN). If it does not, the old
1047        optimizations are restored. It would be nice to do this on a per-group
1048        basis, but at the moment that is not feasible.
1049    
1050    12. In some environments, the output of pcretest -C is CRLF terminated. This
1051        broke RunTest's code that checks for the link size. A single white space
1052        character after the value is now allowed for.
1053    
1054    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
1055        For "fr", it uses the Windows-specific input and output files.
1056    
1057    14. If (*THEN) appeared in a group that was called recursively or as a
1058        subroutine, it did not work as intended. [But see next item.]
1059    
1060    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
1061        pattern fragments (but not containing any | characters). If A and B are
1062        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
1063        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
1064        D. In other words, Perl considers parentheses that do not contain any |
1065        characters to be part of a surrounding alternative, whereas PCRE was
1066        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
1067        differently. PCRE now behaves in the same way as Perl, except in the case
1068        of subroutine/recursion calls such as (?1) which have in any case always
1069        been different (but PCRE had them first :-).
1070    
1071    16. Related to 15 above: Perl does not treat the | in a conditional group as
1072        creating alternatives. Such a group is treated in the same way as an
1073        ordinary group without any | characters when processing (*THEN). PCRE has
1074        been changed to match Perl's behaviour.
1075    
1076    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1077        RunGrepTest script failed.
1078    
1079    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1080        inevitable for groups that contain captures, but it can lead to a lot of
1081        stack use in large patterns. The old behaviour has been restored for atomic
1082        groups that do not contain any capturing parentheses.
1083    
1084    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1085        suppress the check for a minimum subject length at run time. (If it was
1086        given to pcre_exec() or pcre_dfa_exec() it did work.)
1087    
1088    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1089        fail to work when decoding hex characters in data strings in EBCDIC
1090        environments.
1091    
1092    21. It appears that in at least one Mac OS environment, the isxdigit() function
1093        is implemented as a macro that evaluates to its argument more than once,
1094        contravening the C 90 Standard (I haven't checked a later standard). There
1095        was an instance in pcretest which caused it to go wrong when processing
1096        \x{...} escapes in subject strings. The has been rewritten to avoid using
1097        things like p++ in the argument of isxdigit().
1098    
1099    
1100    Version 8.13 16-Aug-2011
1101    ------------------------
1102    
1103    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1104    
1105    2.  Two minor typos in pcre_internal.h have been fixed.
1106    
1107    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1108        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1109        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1110    
1111    4.  There were a number of related bugs in the code for matching backrefences
1112        caselessly in UTF-8 mode when codes for the characters concerned were
1113        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1114        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1115        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1116        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1117        2-byte code at the end of the subject (it thought there wasn't enough data
1118        left).
1119    
1120    5.  Comprehensive information about what went wrong is now returned by
1121        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1122        as the output vector has at least 2 elements. The offset of the start of
1123        the failing character and a reason code are placed in the vector.
1124    
1125    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1126        now returned is for the first byte of the failing character, instead of the
1127        last byte inspected. This is an incompatible change, but I hope it is small
1128        enough not to be a problem. It makes the returned offset consistent with
1129        pcre_exec() and pcre_dfa_exec().
1130    
1131    7.  pcretest now gives a text phrase as well as the error number when
1132        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1133        failure, the offset and reason code are output.
1134    
1135    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1136        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1137        back over a single character (\n). This seems wrong (because it treated the
1138        two characters as a single entity when going forwards), conflicts with the
1139        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1140        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1141        has been changed.
1142    
1143    9.  Some internal refactoring has changed the processing so that the handling
1144        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1145        time (the PCRE_DOTALL option was changed this way some time ago: version
1146        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1147        which was always a bit of a fudge. It also means that there is one less
1148        argument for the match() function, which reduces its stack requirements
1149        slightly. This change also fixes an incompatibility with Perl: the pattern
1150        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1151    
1152    10. More internal refactoring has drastically reduced the number of recursive
1153        calls to match() for possessively repeated groups such as (abc)++ when
1154        using pcre_exec().
1155    
1156    11. While implementing 10, a number of bugs in the handling of groups were
1157        discovered and fixed:
1158    
1159        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1160        (a|)*(?1) gave a compile-time internal error.
1161        ((a|)+)+  did not notice that the outer group could match an empty string.
1162        (^a|^)+   was not marked as anchored.
1163        (.*a|.*)+ was not marked as matching at start or after a newline.
1164    
1165    12. Yet more internal refactoring has removed another argument from the match()
1166        function. Special calls to this function are now indicated by setting a
1167        value in a variable in the "match data" data block.
1168    
1169    13. Be more explicit in pcre_study() instead of relying on "default" for
1170        opcodes that mean there is no starting character; this means that when new
1171        ones are added and accidentally left out of pcre_study(), testing should
1172        pick them up.
1173    
1174    14. The -s option of pcretest has been documented for ages as being an old
1175        synonym of -m (show memory usage). I have changed it to mean "force study
1176        for every regex", that is, assume /S for every regex. This is similar to -i
1177        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1178        using it. It makes it easier to run collections of tests with and without
1179        study enabled, and thereby test pcre_study() more easily. All the standard
1180        tests are now run with and without -s (but some patterns can be marked as
1181        "never study" - see 20 below).
1182    
1183    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1184        restoration of the capturing data to the outer values was not happening
1185        correctly.
1186    
1187    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1188        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1189        pattern had matched an empty string, and so incorrectly returned a no
1190        match.
1191    
1192    17. There was optimizing code for the last branch of non-capturing parentheses,
1193        and also for the obeyed branch of a conditional subexpression, which used
1194        tail recursion to cut down on stack usage. Unfortunately, now that there is
1195        the possibility of (*THEN) occurring in these branches, tail recursion is
1196        no longer possible because the return has to be checked for (*THEN). These
1197        two optimizations have therefore been removed. [But see 8.20/11 above.]
1198    
1199    18. If a pattern containing \R was studied, it was assumed that \R always
1200        matched two bytes, thus causing the minimum subject length to be
1201        incorrectly computed because \R can also match just one byte.
1202    
1203    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1204        was incorrectly computed.
1205    
1206    20. If /S is present twice on a test pattern in pcretest input, it now
1207        *disables* studying, thereby overriding the use of -s on the command line
1208        (see 14 above). This is necessary for one or two tests to keep the output
1209        identical in both cases.
1210    
1211    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1212        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1213    
1214    22. When an atomic group that contained a capturing parenthesis was
1215        successfully matched, but the branch in which it appeared failed, the
1216        capturing was not being forgotten if a higher numbered group was later
1217        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1218        group 1 to "a", when in fact it should be unset. This applied to multi-
1219        branched capturing and non-capturing groups, repeated or not, and also to
1220        positive assertions (capturing in negative assertions does not happen
1221        in PCRE) and also to nested atomic groups.
1222    
1223    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1224        subject after a captured substring, to make it easier to tell which of a
1225        number of identical substrings has been captured.
1226    
1227    24. The way atomic groups are processed by pcre_exec() has been changed so that
1228        if they are repeated, backtracking one repetition now resets captured
1229        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1230        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1231        "aaa". Previously, it would have been "a". As part of this code
1232        refactoring, the way recursive calls are handled has also been changed.
1233    
1234    25. If an assertion condition captured any substrings, they were not passed
1235        back unless some other capturing happened later. For example, if
1236        (?(?=(a))a) was matched against "a", no capturing was returned.
1237    
1238    26. When studying a pattern that contained subroutine calls or assertions,
1239        the code for finding the minimum length of a possible match was handling
1240        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1241        group 1 called group 2 while simultaneously a separate group 2 called group
1242        1). A stack overflow occurred in this case. I have fixed this by limiting
1243        the recursion depth to 10.
1244    
1245    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1246        Fortmann. This supports explicit test numbers on the command line, and has
1247        argument validation and error reporting.
1248    
1249    28. An instance of \X with an unlimited repeat could fail if at any point the
1250        first character it looked at was a mark character.
1251    
1252    29. Some minor code refactoring concerning Unicode properties and scripts
1253        should reduce the stack requirement of match() slightly.
1254    
1255    30. Added the '=' option to pcretest to check the setting of unused capturing
1256        slots at the end of the pattern, which are documented as being -1, but are
1257        not included in the return count.
1258    
1259    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1260        compiled something random. Now it gives a compile-time error (as does
1261        Perl).
1262    
1263    32. A *MARK encountered during the processing of a positive assertion is now
1264        recorded and passed back (compatible with Perl).
1265    
1266    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1267        had alternative anchored branches, the search for a second match in a line
1268        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1269        matched the line "0102" twice. The same bug affected patterns that started
1270        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1271        twice.
1272    
1273    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1274        does, and because of capturing effects, quantifying parenthesized
1275        assertions may at times be useful. Quantifiers are now allowed for
1276        parenthesized assertions.
1277    
1278    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1279    
1280    36. \g was being checked for fancy things in a character class, when it should
1281        just be a literal "g".
1282    
1283    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1284        appearance of a nested POSIX class supersedes an apparent external class.
1285        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1286        unescaped square brackets may also appear as part of class names. For
1287        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1288        more like Perl. (But see 8.20/1 above.)
1289    
1290    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1291        was because it thought it was \N{name}, which is not supported).
1292    
1293    39. Add minix to OS list not supporting the -S option in pcretest.
1294    
1295    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1296        cannot analyze patterns in sufficient detail to catch mutual recursions
1297        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1298        subgroup is called recursively as a subpattern for a second time at the
1299        same position in the subject string. In previous releases this might have
1300        been caught by the recursion limit, or it might have run out of stack.
1301    
1302    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1303        happen only once. PCRE was, however incorrectly giving a compile time error
1304        "recursive call could loop indefinitely" because it cannot analyze the
1305        pattern in sufficient detail. The compile time test no longer happens when
1306        PCRE is compiling a conditional subpattern, but actual runaway loops are
1307        now caught at runtime (see 40 above).
1308    
1309    42. It seems that Perl allows any characters other than a closing parenthesis
1310        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1311        has been changed to be the same.
1312    
1313    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1314        as not to get warnings when autogen.sh is called. Also changed
1315        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1316    
1317    44. To help people who use pcregrep to scan files containing exceedingly long
1318        lines, the following changes have been made:
1319    
1320        (a) The default value of the buffer size parameter has been increased from
1321            8K to 20K. (The actual buffer used is three times this size.)
1322    
1323        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1324            PCRE is built.
1325    
1326        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1327            to be set at run time.
1328    
1329        (d) Numerical values in pcregrep options can be followed by K or M, for
1330            example --buffer-size=50K.
1331    
1332        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1333            given and the return code is set to 2.
1334    
1335    45. Add a pointer to the latest mark to the callout data block.
1336    
1337    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1338        partial match of an empty string instead of no match. This was specific to
1339        the use of ".".
1340    
1341    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1342        complete match instead of a partial match. This bug was dependent on both
1343        the PCRE_UTF8 and PCRE_DOTALL options being set.
1344    
1345    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1346        starting byte set, because \b was not being ignored.
1347    
1348    
1349    Version 8.12 15-Jan-2011
1350    ------------------------
1351    
1352    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1353        checks for such things as part of the documentation building process.
1354    
1355    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1356        --match-limit and --recursion-limit options (added for 8.11). In
1357        particular, this made one of the standard tests fail. (The integer value
1358        went into the wrong half of a long int.)
1359    
1360    3.  If the --colour option was given to pcregrep with -v (invert match), it
1361        did strange things, either producing crazy output, or crashing. It should,
1362        of course, ignore a request for colour when reporting lines that do not
1363        match.
1364    
1365    4.  Another pcregrep bug caused similar problems if --colour was specified with
1366        -M (multiline) and the pattern match finished with a line ending.
1367    
1368    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1369        matched in multiline mode, the following line was shown as part of the
1370        match. This seems wrong, so I have changed it.
1371    
1372    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1373        the check for further matches in the same line (so they could be coloured)
1374        to overrun the end of the current line. If another match was found, it was
1375        incorrectly shown (and then shown again when found in the next line).
1376    
1377    7.  If pcregrep was compiled under Windows, there was a reference to the
1378        function pcregrep_exit() before it was defined. I am assuming this was
1379        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1380        reported by a user. I've moved the definition above the reference.
1381    
1382    
1383    Version 8.11 10-Dec-2010
1384    ------------------------
1385    
1386    1.  (*THEN) was not working properly if there were untried alternatives prior
1387        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1388        backtracked to try for "b" instead of moving to the next alternative branch
1389        at the same level (in this case, to look for "c"). The Perl documentation
1390        is clear that when (*THEN) is backtracked onto, it goes to the "next
1391        alternative in the innermost enclosing group".
1392    
1393    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1394        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1395        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1396        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1397        (*THEN).
1398    
1399    3.  If \s appeared in a character class, it removed the VT character from
1400        the class, even if it had been included by some previous item, for example
1401        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1402        of \s, but is part of the POSIX "space" class.)
1403    
1404    4.  A partial match never returns an empty string (because you can always
1405        match an empty string at the end of the subject); however the checking for
1406        an empty string was starting at the "start of match" point. This has been
1407        changed to the "earliest inspected character" point, because the returned
1408        data for a partial match starts at this character. This means that, for
1409        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1410        (previously it gave "no match").
1411    
1412    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1413        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1414        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1415        has an implication that the given string is incomplete (because a partial
1416        match is preferred over a full match). For this reason, these items now
1417        give a partial match in this situation. [Aside: previously, the one case
1418        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1419        match rather than a full match, which was wrong by the old rules, but is
1420        now correct.]
1421    
1422    6.  There was a bug in the handling of #-introduced comments, recognized when
1423        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1424        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1425        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1426        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1427        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1428        places in pcre_compile().
1429    
1430    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1431        comments when looking ahead for named forward references to subpatterns,
1432        the only newline sequence it recognized was NL. It now handles newlines
1433        according to the set newline convention.
1434    
1435    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1436        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1437        cater for a lack of strerror(). These oversights have been fixed.
1438    
1439    9.  Added --match-limit and --recursion-limit to pcregrep.
1440    
1441    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1442    
1443    11. When the -o option was used, pcregrep was setting a return code of 1, even
1444        when matches were found, and --line-buffered was not being honoured.
1445    
1446    12. Added an optional parentheses number to the -o and --only-matching options
1447        of pcregrep.
1448    
1449    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1450        can match an empty string. The code to do it in pcretest and pcredemo
1451        needed fixing:
1452    
1453        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1454            only one byte after an empty string match just before CRLF (this case
1455            just got forgotten; "any" and "anycrlf" were OK).
1456    
1457        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1458            mode when an empty string match preceded an ASCII character followed by
1459            a non-ASCII character. (The code for advancing by one character rather
1460            than one byte was nonsense.)
1461    
1462        (c) The pcredemo.c sample program did not have any code at all to handle
1463            the cases when CRLF is a valid newline sequence.
1464    
1465    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1466        as a starting offset was within the subject string. There is now a new
1467        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1468        negative or greater than the length of the string. In order to test this,
1469        pcretest is extended to allow the setting of negative starting offsets.
1470    
1471    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1472        starting offset points to the beginning of a UTF-8 character was
1473        unnecessarily clumsy. I tidied it up.
1474    
1475    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1476        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1477    
1478    17. Nobody had reported that the --include_dir option, which was added in
1479        release 7.7 should have been called --include-dir (hyphen, not underscore)
1480        for compatibility with GNU grep. I have changed it to --include-dir, but
1481        left --include_dir as an undocumented synonym, and the same for
1482        --exclude-dir, though that is not available in GNU grep, at least as of
1483        release 2.5.4.
1484    
1485    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1486        characters from a string of bytes) have been redefined so as not to use
1487        loops, in order to improve performance in some environments. At the same
1488        time, I abstracted some of the common code into auxiliary macros to save
1489        repetition (this should not affect the compiled code).
1490    
1491    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1492        compile-time error is now given if \c is not followed by an ASCII
1493        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1494        different, and any byte value is allowed.)
1495    
1496    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1497        START_OPTIMIZE option, which is now allowed at compile time - but just
1498        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1499        to pcregrep and other applications that have no direct access to PCRE
1500        options. The new /Y option in pcretest sets this option when calling
1501        pcre_compile().
1502    
1503    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1504        back references. Groups containing recursive back references were forced to
1505        be atomic by that change, but in the case of named groups, the amount of
1506        memory required was incorrectly computed, leading to "Failed: internal
1507        error: code overflow". This has been fixed.
1508    
1509    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1510        pcretest.c, to avoid build problems in some Borland environments.
1511    
1512    
1513    Version 8.10 25-Jun-2010
1514    ------------------------
1515    
1516    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1517        THEN.
1518    
1519    2.  (*ACCEPT) was not working when inside an atomic group.
1520    
1521    3.  Inside a character class, \B is treated as a literal by default, but
1522        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1523        causes the error). The code is unchanged, but I tidied the documentation.
1524    
1525    4.  Inside a character class, PCRE always treated \R and \X as literals,
1526        whereas Perl faults them if its -w option is set. I have changed PCRE so
1527        that it faults them when PCRE_EXTRA is set.
1528    
1529    5.  Added support for \N, which always matches any character other than
1530        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1531    
1532    6.  When compiling pcregrep with newer versions of gcc which may have
1533        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1534        declared with attribute warn_unused_result" were given. Just casting the
1535        result to (void) does not stop the warnings; a more elaborate fudge is
1536        needed. I've used a macro to implement this.
1537    
1538    7.  Minor change to pcretest.c to avoid a compiler warning.
1539    
1540    8.  Added four artifical Unicode properties to help with an option to make
1541        \s etc use properties (see next item). The new properties are: Xan
1542        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1543    
1544    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1545        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1546        this option. Modified pcretest to add /W to test this facility. Added
1547        REG_UCP to make it available via the POSIX interface.
1548    
1549    10. Added --line-buffered to pcregrep.
1550    
1551    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1552        studied, and the match started with a letter with a code point greater than
1553        127 whose first byte was different to the first byte of the other case of
1554        the letter, the other case of this starting letter was not recognized
1555        (#976).
1556    
1557    12. If a pattern that was studied started with a repeated Unicode property
1558        test, for example, \p{Nd}+, there was the theoretical possibility of
1559        setting up an incorrect bitmap of starting bytes, but fortunately it could
1560        not have actually happened in practice until change 8 above was made (it
1561        added property types that matched character-matching opcodes).
1562    
1563    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1564        possible starting bytes for non-anchored patterns.
1565    
1566    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1567        \R, and also a number of cases that involve Unicode properties, both
1568        explicit and implicit when PCRE_UCP is set.
1569    
1570    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1571        input, it could crash or give wrong results if characters with values
1572        greater than 0xc0 were present in the subject string. (Detail: it assumed
1573        UTF-8 input when processing these items.)
1574    
1575    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1576        size_t is 64-bit (#991).
1577    
1578    17. Added a check for running out of memory when PCRE is compiled with
1579        --disable-stack-for-recursion (#990).
1580    
1581    18. If the last data line in a file for pcretest does not have a newline on
1582        the end, a newline was missing in the output.
1583    
1584    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1585        less than 128) in its various bitmaps. However, there is a facility for
1586        generating tables according to the current locale when PCRE is compiled. It
1587        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1588        characters, are recognized by isspace() and therefore were getting set in
1589        these tables, and indeed these tables seem to approximate to ISO 8859. This
1590        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1591        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1592        which of course cannot start UTF-8 characters. I have changed the code so
1593        that only real ASCII characters (less than 128) and the correct starting
1594        bytes for UTF-8 encodings are set for characters greater than 127 when in
1595        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1596        altogether.)
1597    
1598    20. Added the /T option to pcretest so as to be able to run tests with non-
1599        standard character tables, thus making it possible to include the tests
1600        used for 19 above in the standard set of tests.
1601    
1602    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1603        reference to a subpattern the other side of a comment that contains an
1604        opening parenthesis caused either an internal compiling error, or a
1605        reference to the wrong subpattern.
1606    
1607    
1608    Version 8.02 19-Mar-2010
1609    ------------------------
1610    
1611    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1612    
1613    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1614        configured.
1615    
1616    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1617        original author of that file, following a query about its status.
1618    
1619    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1620        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1621    
1622    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1623        quantifier applied to a forward-referencing subroutine call, could compile
1624        incorrect code or give the error "internal error: previously-checked
1625        referenced subpattern not found".
1626    
1627    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1628        variables to point to external functions. For these systems, therefore,
1629        pcre_malloc etc. are now initialized to local functions that call the
1630        relevant global functions.
1631    
1632    7.  There were two entries missing in the vectors called coptable and poptable
1633        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1634        I've fixed the data, and added a kludgy way of testing at compile time that
1635        the lengths are correct (equal to the number of opcodes).
1636    
1637    8.  Following on from 7, I added a similar kludge to check the length of the
1638        eint vector in pcreposix.c.
1639    
1640    9.  Error texts for pcre_compile() are held as one long string to avoid too
1641        much relocation at load time. To find a text, the string is searched,
1642        counting zeros. There was no check for running off the end of the string,
1643        which could happen if a new error number was added without updating the
1644        string.
1645    
1646    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1647    
1648    11. \K was not working if it appeared in an atomic group or in a group that
1649        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1650        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1651        the assertion is positive, but not if it is negative.
1652    
1653    12. Change 11 fortuitously reduced the size of the stack frame used in the
1654        "match()" function of pcre_exec.c by one pointer. Forthcoming
1655        implementation of support for (*MARK) will need an extra pointer on the
1656        stack; I have reserved it now, so that the stack frame size does not
1657        decrease.
1658    
1659    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1660        item in branch that calls a recursion is a subroutine call - as in the
1661        second branch in the above example - was incorrectly given the compile-
1662        time error "recursive call could loop indefinitely" because pcre_compile()
1663        was not correctly checking the subroutine for matching a non-empty string.
1664    
1665    14. The checks for overrunning compiling workspace could trigger after an
1666        overrun had occurred. This is a "should never occur" error, but it can be
1667        triggered by pathological patterns such as hundreds of nested parentheses.
1668        The checks now trigger 100 bytes before the end of the workspace.
1669    
1670    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1671    
1672    
1673    Version 8.01 19-Jan-2010
1674    ------------------------
1675    
1676    1.  If a pattern contained a conditional subpattern with only one branch (in
1677        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1678        computed the wrong minimum data length (which is of course zero for such
1679        subpatterns). This could cause incorrect "no match" results.
1680    
1681    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1682        the pattern is reset in the first branch, pcre_compile() failed with
1683        "internal error: code overflow at offset...". This happened only when
1684        the reset was to the original external option setting. (An optimization
1685        abstracts leading options settings into an external setting, which was the
1686        cause of this.)
1687    
1688    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1689        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1690        assertion pattern did not match (meaning that the assertion was true), it
1691        was incorrectly treated as false if the SKIP had been reached during the
1692        matching. This also applied to assertions used as conditions.
1693    
1694    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1695        assertion subpattern, including such a pattern used as a condition,
1696        unpredictable results occurred, instead of the error return
1697        PCRE_ERROR_DFA_UITEM.
1698    
1699    5.  The C++ GlobalReplace function was not working like Perl for the special
1700        situation when an empty string is matched. It now does the fancy magic
1701        stuff that is necessary.
1702    
1703    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1704        removed. (These were left over from very, very early versions of PCRE.)
1705    
1706    7.  Some cosmetic changes to the code to make life easier when compiling it
1707        as part of something else:
1708    
1709        (a) Change DEBUG to PCRE_DEBUG.
1710    
1711        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1712            called "current" as "current_branch", to prevent a collision with the
1713            Linux macro when compiled as a kernel module.
1714    
1715        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1716            prevent a collision with the Linux macro when compiled as a kernel
1717            module.
1718    
1719    8.  In pcre_compile() there are some checks for integer overflows that used to
1720        cast potentially large values to (double). This has been changed to that
1721        when building, a check for int64_t is made, and if it is found, it is used
1722        instead, thus avoiding the use of floating point arithmetic. (There is no
1723        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1724        double.
1725    
1726    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1727        2005 (difference between two addresses compared to an unsigned value).
1728    
1729    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1730        custom one, because of the following reported problem in Windows:
1731    
1732          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1733              under Win32.
1734          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1735              therefore missing the function definition.
1736          - The compiler thus generates a "C" signature for the test function.
1737          - The linker fails to find the "C" function.
1738          - PCRE fails to configure if asked to do so against libbz2.
1739    
1740    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1741        messages were output:
1742    
1743          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1744          rerunning libtoolize, to keep the correct libtool macros in-tree.
1745          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1746    
1747        I have done both of these things.
1748    
1749    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1750        most of the time, it *can* run out if it is given a pattern that contains a
1751        runaway infinite recursion. I updated the discussion in the pcrestack man
1752        page.
1753    
1754    13. Now that we have gone to the x.xx style of version numbers, the minor
1755        version may start with zero. Using 08 or 09 is a bad idea because users
1756        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1757        interpreted as invalid octal numbers. I've updated the previous comment in
1758        configure.ac, and also added a check that gives an error if 08 or 09 are
1759        used.
1760    
1761    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1762        causing partial matching to fail when the end of the subject matched \W
1763        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1764    
1765    15. There were some discrepancies between the declarations in pcre_internal.h
1766        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1767        their definitions. The declarations used "const uschar *" and the
1768        definitions used USPTR. Even though USPTR is normally defined as "const
1769        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1770        reported that: "This difference in casting confuses some C++ compilers, for
1771        example, SunCC recognizes above declarations as different functions and
1772        generates broken code for hbpcre." I have changed the declarations to use
1773        USPTR.
1774    
1775    16. GNU libtool is named differently on some systems. The autogen.sh script now
1776        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1777        (FreeBSD).
1778    
1779    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1780        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1781        comment: "Figure out how to create a longlong from a string: strtoll and
1782        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1783        instance, but it only takes 2 args instead of 3!"
1784    
1785    18. A subtle bug concerned with back references has been fixed by a change of
1786        specification, with a corresponding code fix. A pattern such as
1787        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1788        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1789        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1790        same bug. Such groups have to be quantified to be useful, or contained
1791        inside another quantified group. (If there's no repetition, the reference
1792        can never match.) The problem arises because, having left the group and
1793        moved on to the rest of the pattern, a later failure that backtracks into
1794        the group uses the captured value from the final iteration of the group
1795        rather than the correct earlier one. I have fixed this in PCRE by forcing
1796        any group that contains a reference to itself to be an atomic group; that
1797        is, there cannot be any backtracking into it once it has completed. This is
1798        similar to recursive and subroutine calls.
1799    
1800    
1801    Version 8.00 19-Oct-09
1802    ----------------------
1803    
1804    1.  The table for translating pcre_compile() error codes into POSIX error codes
1805        was out-of-date, and there was no check on the pcre_compile() error code
1806        being within the table. This could lead to an OK return being given in
1807        error.
1808    
1809    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1810        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1811        in a Windows environment.
1812    
1813    3.  The pcregrep --count option prints the count for each file even when it is
1814        zero, as does GNU grep. However, pcregrep was also printing all files when
1815        --files-with-matches was added. Now, when both options are given, it prints
1816        counts only for those files that have at least one match. (GNU grep just
1817        prints the file name in this circumstance, but including the count seems
1818        more useful - otherwise, why use --count?) Also ensured that the
1819        combination -clh just lists non-zero counts, with no names.
1820    
1821    4.  The long form of the pcregrep -F option was incorrectly implemented as
1822        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1823        but it seems right to fix it, and I didn't think it was worth preserving
1824        the old behaviour.
1825    
1826    5.  The command line items --regex=pattern and --regexp=pattern were not
1827        recognized by pcregrep, which required --regex pattern or --regexp pattern
1828        (with a space rather than an '='). The man page documented the '=' forms,
1829        which are compatible with GNU grep; these now work.
1830    
1831    6.  No libpcreposix.pc file was created for pkg-config; there was just
1832        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1833    
1834    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1835        when UCP support is not needed, by modifying the Python script that
1836        generates it from Unicode data files. This should not matter if the module
1837        is correctly used as a library, but I received one complaint about 50K of
1838        unwanted data. My guess is that the person linked everything into his
1839        program rather than using a library. Anyway, it does no harm.
1840    
1841    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1842        was a minimum greater than 1 for a wide character in a possessive
1843        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1844        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1845        character. Chaos in the form of incorrect output or a compiling loop could
1846        result.
1847    
1848    9.  The restrictions on what a pattern can contain when partial matching is
1849        requested for pcre_exec() have been removed. All patterns can now be
1850        partially matched by this function. In addition, if there are at least two
1851        slots in the offset vector, the offset of the earliest inspected character
1852        for the match and the offset of the end of the subject are set in them when
1853        PCRE_ERROR_PARTIAL is returned.
1854    
1855    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1856        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1857        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1858        and may be more useful for multi-segment matching.
1859    
1860    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1861        used to be given if ever the end of the subject was reached; now it is
1862        given only if matching could not proceed because another character was
1863        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1864        string "Z", which now yields "no match" instead of "partial match". In the
1865        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1866        final character ended with (*FAIL).
1867    
1868    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1869        if the pattern had a "must contain" character that was already found in the
1870        earlier partial match, unless partial matching was again requested. For
1871        example, with the pattern /dog.(body)?/, the "must contain" character is
1872        "g". If the first part-match was for the string "dog", restarting with
1873        "sbody" failed. This bug has been fixed.
1874    
1875    13. The string returned by pcre_dfa_exec() after a partial match has been
1876        changed so that it starts at the first inspected character rather than the
1877        first character of the match. This makes a difference only if the pattern
1878        starts with a lookbehind assertion or \b or \B (\K is not supported by
1879        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1880        matching functions compatible, and I think it's the right thing to do.
1881    
1882    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1883        so that the demonstration program is easily available in environments where
1884        PCRE has not been installed from source.
1885    
1886    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1887        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1888        library.
1889    
1890    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1891        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1892        is not the first non-POSIX option to be added. Clearly some people find
1893        these options useful.
1894    
1895    17. If a caller to the POSIX matching function regexec() passes a non-zero
1896        value for nmatch with a NULL value for pmatch, the value of
1897        nmatch is forced to zero.
1898    
1899    18. RunGrepTest did not have a test for the availability of the -u option of
1900        the diff command, as RunTest does. It now checks in the same way as
1901        RunTest, and also checks for the -b option.
1902    
1903    19. If an odd number of negated classes containing just a single character
1904        interposed, within parentheses, between a forward reference to a named
1905        subpattern and the definition of the subpattern, compilation crashed with
1906        an internal error, complaining that it could not find the referenced
1907        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1908        [The bug was that it was starting one character too far in when skipping
1909        over the character class, thus treating the ] as data rather than
1910        terminating the class. This meant it could skip too much.]
1911    
1912    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1913        /g option in pcretest when the pattern contains \K, which makes it possible
1914        to have an empty string match not at the start, even when the pattern is
1915        anchored. Updated pcretest and pcredemo to use this option.
1916    
1917    21. If the maximum number of capturing subpatterns in a recursion was greater
1918        than the maximum at the outer level, the higher number was returned, but
1919        with unset values at the outer level. The correct (outer level) value is
1920        now given.
1921    
1922    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1923        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1924        make it do so. The string so far is captured, making this feature
1925        compatible with Perl.
1926    
1927    23. The tests have been re-organized, adding tests 11 and 12, to make it
1928        possible to check the Perl 5.10 features against Perl 5.10.
1929    
1930    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1931        pattern matches a fixed length string. PCRE did not allow this; now it
1932        does. Neither allows recursion.
1933    
1934    25. I finally figured out how to implement a request to provide the minimum
1935        length of subject string that was needed in order to match a given pattern.
1936        (It was back references and recursion that I had previously got hung up
1937        on.) This code has now been added to pcre_study(); it finds a lower bound
1938        to the length of subject needed. It is not necessarily the greatest lower
1939        bound, but using it to avoid searching strings that are too short does give
1940        some useful speed-ups. The value is available to calling programs via
1941        pcre_fullinfo().
1942    
1943    26. While implementing 25, I discovered to my embarrassment that pcretest had
1944        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1945        study optimizations had never been tested with that matching function.
1946        Oops. What is worse, even when it was passed study data, there was a bug in
1947        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1948        were also very few tests of studied patterns with pcre_dfa_exec().
1949    
1950    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1951        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1952        on the other side of the coin, they are no longer allowed to have different
1953        names, because these cannot be distinguished in PCRE, and this has caused
1954        confusion. (This is a difference from Perl.)
1955    
1956    28. When duplicate subpattern names are present (necessarily with different
1957        numbers, as required by 27 above), and a test is made by name in a
1958        conditional pattern, either for a subpattern having been matched, or for
1959        recursion in such a pattern, all the associated numbered subpatterns are
1960        tested, and the overall condition is true if the condition is true for any
1961        one of them. This is the way Perl works, and is also more like the way
1962        testing by number works.
1963    
1964    
1965    Version 7.9 11-Apr-09
1966    ---------------------
1967    
1968    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1969        (pcretest), all targets were linked against these libraries. This included
1970        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1971        libraries. This caused unwanted dependencies to be created. This problem
1972        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1973        pcretest is linked with readline.
1974    
1975    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1976        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1977        moved outside it again, because FALSE and TRUE are already defined in AIX,
1978        but BOOL is not.
1979    
1980    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1981        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1982    
1983    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1984        hyphens) following file names and line numbers when outputting matching
1985        lines. This is not true; no spaces are inserted. I have also clarified the
1986        wording for the --colour (or --color) option.
1987    
1988    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1989        was not coloured; this is different to GNU grep, so I have changed it to be
1990        the same.
1991    
1992    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1993        each matching line was coloured. Now it goes on to look for further matches
1994        of any of the test patterns, which is the same behaviour as GNU grep.
1995    
1996    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1997        doesn't make sense to accept an empty string match in pcregrep, so I have
1998        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1999        seems to be how GNU grep behaves. [But see later change 40 for release
2000        8.33.]
2001    
2002    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
2003        start or after a newline", because the conditional assertion was not being
2004        correctly handled. The rule now is that both the assertion and what follows
2005        in the first alternative must satisfy the test.
2006    
2007    9.  If auto-callout was enabled in a pattern with a conditional group whose
2008        condition was an assertion, PCRE could crash during matching, both with
2009        pcre_exec() and pcre_dfa_exec().
2010    
2011    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
2012        used for matching.
2013    
2014    11. Unicode property support in character classes was not working for
2015        characters (bytes) greater than 127 when not in UTF-8 mode.
2016    
2017    12. Added the -M command line option to pcretest.
2018    
2019    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
2020    
2021    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
2022    
2023    16. Added comments and documentation about mis-use of no_arg in the C++
2024        wrapper.
2025    
2026    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
2027        from Martin Jerabek that uses macro names for all relevant character and
2028        string constants.
2029    
2030    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
2031        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
2032        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
2033        these, but not everybody uses configure.
2034    
2035    19. A conditional group that had only one branch was not being correctly
2036        recognized as an item that could match an empty string. This meant that an
2037        enclosing group might also not be so recognized, causing infinite looping
2038        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
2039        with the subject "ab", where knowledge that the repeated group can match
2040        nothing is needed in order to break the loop.
2041    
2042    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
2043        exec(), but without supplying a callout function, matching went wrong.
2044    
2045    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
2046        leak if the size of the offset vector was greater than 30. When the vector
2047        is smaller, the saved offsets during recursion go onto a local stack
2048        vector, but for larger vectors malloc() is used. It was failing to free
2049        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
2050        error, in fact).
2051    
2052    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
2053        heapframe that is used only when UTF-8 support is enabled. This caused no
2054        problem, but was untidy.
2055    
2056    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
2057        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
2058        included within another project.
2059    
2060    24. Steven Van Ingelgem's patches to add more options to the CMake support,
2061        slightly modified by me:
2062    
2063          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
2064              not building pcregrep.
2065    
2066          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
2067              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
2068    
2069    25. Forward references, both numeric and by name, in patterns that made use of
2070        duplicate group numbers, could behave incorrectly or give incorrect errors,
2071        because when scanning forward to find the reference group, PCRE was not
2072        taking into account the duplicate group numbers. A pattern such as
2073        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
2074    
2075    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2076        the feature of a custom pointer more persuasive (as requested by a user).
2077    
2078    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2079        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2080        pre-defined.
2081    
2082    28. Added support for (*UTF8) at the start of a pattern.
2083    
2084    29. Arrange for flags added by the "release type" setting in CMake to be shown
2085        in the configuration summary.
2086    
2087    
2088    Version 7.8 05-Sep-08
2089    ---------------------
2090    
2091    1.  Replaced UCP searching code with optimized version as implemented for Ad
2092        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2093        stage table and inline lookup instead of a function, giving speed ups of 2
2094        to 5 times on some simple patterns that I tested. Permission was given to
2095        distribute the MultiStage2.py script that generates the tables (it's not in
2096        the tarball, but is in the Subversion repository).
2097    
2098    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2099        scripts.
2100    
2101    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2102        a group with a zero qualifier. The result of the study could be incorrect,
2103        or the function might crash, depending on the pattern.
2104    
2105    4.  Caseless matching was not working for non-ASCII characters in back
2106        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2107        It now works when Unicode Property Support is available.
2108    
2109    5.  In pcretest, an escape such as \x{de} in the data was always generating
2110        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2111        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2112        truncation.
2113    
2114    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2115    
2116    7.  Added two (int) casts to pcregrep when printing the difference of two
2117        pointers, in case they are 64-bit values.
2118    
2119    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2120        test 2 if it fails.
2121    
2122    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2123        and a #define of that name to empty if it is not externally set. This is to
2124        allow users of MSVC to set it if necessary.
2125    
2126    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2127        the convenience functions in the pcre_get.c source file.
2128    
2129    11. An option change at the start of a pattern that had top-level alternatives
2130        could cause overwriting and/or a crash. This command provoked a crash in
2131        some environments:
2132    
2133          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2134    
2135        This potential security problem was recorded as CVE-2008-2371.
2136    
2137    12. For a pattern where the match had to start at the beginning or immediately
2138        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2139        pcre_dfa_exec() could read past the end of the passed subject if there was
2140        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2141        pcretest so that it places the subject at the end of its malloc-ed buffer.
2142    
2143    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2144        exec() might read past the end of the data buffer in UTF-8 mode.
2145    
2146    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2147        the data contained the byte 0x85 as part of a UTF-8 character within its
2148        first line. This applied both to normal and DFA matching.
2149    
2150    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2151        /^[^d]*?$/8 failed to match "abc".
2152    
2153    16. Added a missing copyright notice to pcrecpp_internal.h.
2154    
2155    17. Make it more clear in the documentation that values returned from
2156        pcre_exec() in ovector are byte offsets, not character counts.
2157    
2158    18. Tidied a few places to stop certain compilers from issuing warnings.
2159    
2160    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2161        supplied by Stefan Weber. I made a further small update for 7.8 because
2162        there is a change of source arrangements: the pcre_searchfuncs.c module is
2163        replaced by pcre_ucd.c.
2164    
2165    
2166    Version 7.7 07-May-08
2167    ---------------------
2168    
2169    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2170        a string to a long long, pretend we don't even have a long long." This is
2171        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2172    
2173    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2174        pre-7.6 versions, which defined a global no_arg variable instead of putting
2175        it in the RE class. (See also #8 below.)
2176    
2177    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2178        Lopes.
2179    
2180    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2181    
2182        (1) The include/exclude patterns were being applied to the whole pathnames
2183            of files, instead of just to the final components.
2184    
2185        (2) If there was more than one level of directory, the subdirectories were
2186            skipped unless they satisfied the include/exclude conditions. This is
2187            inconsistent with GNU grep (and could even be seen as contrary to the
2188            pcregrep specification - which I improved to make it absolutely clear).
2189            The action now is always to scan all levels of directory, and just
2190            apply the include/exclude patterns to regular files.
2191    
2192    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2193        --exclude_dir in the tests to avoid scanning .svn directories.
2194    
2195    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2196        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2197        doesn't support NULs in patterns.
2198    
2199    7.  Added some missing "const"s to declarations of static tables in
2200        pcre_compile.c and pcre_dfa_exec.c.
2201    
2202    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2203        caused by fix #2  above. (Subsequently also a second patch to fix the
2204        first patch. And a third patch - this was a messy problem.)
2205    
2206    9.  Applied Craig's patch to remove the use of push_back().
2207    
2208    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2209        matching function regexec().
2210    
2211    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2212        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2213        references. PCRE supports relative numbers with this syntax (I don't think
2214        Oniguruma does).
2215    
2216    12. Previously, a group with a zero repeat such as (...){0} was completely
2217        omitted from the compiled regex. However, this means that if the group
2218        was called as a subroutine from elsewhere in the pattern, things went wrong
2219        (an internal error was given). Such groups are now left in the compiled
2220        pattern, with a new opcode that causes them to be skipped at execution
2221        time.
2222    
2223    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2224        to the way PCRE behaves:
2225    
2226        (a) A lone ] character is dis-allowed (Perl treats it as data).
2227    
2228        (b) A back reference to an unmatched subpattern matches an empty string
2229            (Perl fails the current match path).
2230    
2231        (c) A data ] in a character class must be notated as \] because if the
2232            first data character in a class is ], it defines an empty class. (In
2233            Perl it is not possible to have an empty class.) The empty class []
2234            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2235            The negative empty class [^] matches any one character, independently
2236            of the DOTALL setting.
2237    
2238    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2239        non-existent subpattern following a character class starting with ']' and
2240        containing () gave an internal compiling error instead of "reference to
2241        non-existent subpattern". Fortunately, when the pattern did exist, the
2242        compiled code was correct. (When scanning forwards to check for the
2243        existence of the subpattern, it was treating the data ']' as terminating
2244        the class, so got the count wrong. When actually compiling, the reference
2245        was subsequently set up correctly.)
2246    
2247    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2248        it was being rejected as not supported by pcre_dfa_exec(), even though
2249        other assertions are supported. I have made pcre_dfa_exec() support
2250        (*FAIL).
2251    
2252    16. The implementation of 13c above involved the invention of a new opcode,
2253        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2254        cannot be changed at match time, I realized I could make a small
2255        improvement to matching performance by compiling OP_ALLANY instead of
2256        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2257        on the OP_ANY path.
2258    
2259    17. Compiling pcretest on Windows with readline support failed without the
2260        following two fixes: (1) Make the unistd.h include conditional on
2261        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2262    
2263    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2264        ncurses library to be included for pcretest when ReadLine support is
2265        requested, but also to allow for it to be overridden. This patch came from
2266        Daniel Bergström.
2267    
2268    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2269        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2270        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2271        spotting this.
2272    
2273    
2274    Version 7.6 28-Jan-08
2275    ---------------------
2276    
2277    1.  A character class containing a very large number of characters with
2278        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2279        overflow.
2280    
2281    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2282        HAVE_LONG_LONG is not defined.
2283    
2284    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2285        bring them up to date and include new features. This patch includes:
2286    
2287        - Fixed PH's badly added libz and libbz2 support.
2288        - Fixed a problem with static linking.
2289        - Added pcredemo. [But later removed - see 7 below.]
2290        - Fixed dftables problem and added an option.
2291        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2292            HAVE_LONG_LONG.
2293        - Added readline support for pcretest.
2294        - Added an listing of the option settings after cmake has run.
2295    
2296    4.  A user submitted a patch to Makefile that makes it easy to create
2297        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2298        Makefile.am that cause it to include this special target, without
2299        affecting anything else. Note that the same mingw target plus all
2300        the other distribution libraries and programs are now supported
2301        when configuring with CMake (see 6 below) instead of with
2302        Configure/Make.
2303    
2304    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2305        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2306        exported in the Windows port". It has not yet been confirmed that the patch
2307        solves the problem, but it does no harm.
2308    
2309    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2310        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2311        with CMake, and also correct the comment about stack recursion.
2312    
2313    7.  Remove the automatic building of pcredemo from the ./configure system and
2314        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2315        of a program that users should build themselves after PCRE is installed, so
2316        building it automatically is not really right. What is more, it gave
2317        trouble in some build environments.
2318    
2319    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2320    
2321    
2322    Version 7.5 10-Jan-08
2323    ---------------------
2324    
2325    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2326        values in parens when parsing an RE using the C++ wrapper."
2327    
2328    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2329        Characters greater than 255 were excluded from the class instead of being
2330        included.
2331    
2332    3.  The same bug as (2) above applied to negated POSIX classes such as
2333        [:^space:].
2334    
2335    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2336        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2337        I have changed it.
2338    
2339    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2340        first named subpattern) and a construct such as (?&a) would reference the
2341        first named subpattern whose name started with "a" (in other words, the
2342        length check was missing). Both these problems are fixed. "Subpattern name
2343        expected" is now given for (?&) (a zero-length name), and this patch also
2344        makes it give the same error for \k'' (previously it complained that that
2345        was a reference to a non-existent subpattern).
2346    
2347    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2348        this is right because (?- can be followed by option settings as well as by
2349        digits. I have, however, made the messages clearer.
2350    
2351    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2352        than the number used in the conditional) now cause a compile-time error.
2353        This is actually not compatible with Perl, which accepts such patterns, but
2354        treats the conditional as always being FALSE (as PCRE used to), but it
2355        seems to me that giving a diagnostic is better.
2356    
2357    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2358        and messages.
2359    
2360    9.  Fix two occurrences of "backslash" in comments that should have been
2361        "backspace".
2362    
2363    10. Remove two redundant lines of code that can never be obeyed (their function
2364        was moved elsewhere).
2365    
2366    11. The program that makes PCRE's Unicode character property table had a bug
2367        which caused it to generate incorrect table entries for sequences of
2368        characters that have the same character type, but are in different scripts.
2369        It amalgamated them into a single range, with the script of the first of
2370        them. In other words, some characters were in the wrong script. There were
2371        thirteen such cases, affecting characters in the following ranges:
2372    
2373          U+002b0 - U+002c1
2374          U+0060c - U+0060d
2375          U+0061e - U+00612
2376          U+0064b - U+0065e
2377          U+0074d - U+0076d
2378          U+01800 - U+01805
2379          U+01d00 - U+01d77
2380          U+01d9b - U+01dbf
2381          U+0200b - U+0200f
2382          U+030fc - U+030fe
2383          U+03260 - U+0327f
2384          U+0fb46 - U+0fbb1
2385          U+10450 - U+1049d
2386    
2387    12. The -o option (show only the matching part of a line) for pcregrep was not
2388        compatible with GNU grep in that, if there was more than one match in a
2389        line, it showed only the first of them. It now behaves in the same way as
2390        GNU grep.
2391    
2392    13. If the -o and -v options were combined for pcregrep, it printed a blank
2393        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2394        does the same. The return code can be used to tell if there were any
2395        non-matching lines.
2396    
2397    14. Added --file-offsets and --line-offsets to pcregrep.
2398    
2399    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2400        infinitely looping recursion. The bug was that positive lookaheads were not
2401        being skipped when checking for a possible empty match (negative lookaheads
2402        and both kinds of lookbehind were skipped).
2403    
2404    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2405        inclusion of <windows.h> to before rather than after the definition of
2406        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2407    
2408    17. Specifying a possessive quantifier with a specific limit for a Unicode
2409        character property caused pcre_compile() to compile bad code, which led at
2410        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2411        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2412        caused the error; without that there was no problem.
2413    
2414    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2415    
2416    19. Added --enable-pcretest-libreadline.
2417    
2418    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2419        RE::GlobalReplace(). As a result, the number of replacements returned was
2420        double what it should be. I removed one of the increments, but Craig sent a
2421        later patch that removed the other one (the right fix) and added unit tests
2422        that check the return values (which was not done before).
2423    
2424    21. Several CMake things:
2425    
2426        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2427            the names libpcre and libpcreposix, not just pcre and pcreposix.
2428    
2429        (2) The above change means that pcretest and pcregrep are now correctly
2430            linked with the newly-built libraries, not previously installed ones.
2431    
2432        (3) Added PCRE_SUPPORT_LIBREADLINE, PCRE_SUPPORT_LIBZ, PCRE_SUPPORT_LIBBZ2.
2433    
2434    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2435        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2436        UTF-8 newline character). The key issue is that the pattern starts .*;
2437        this means that the match must be either at the beginning, or after a
2438        newline. The bug was in the code for advancing after a failed match and
2439        checking that the new position followed a newline. It was not taking
2440        account of UTF-8 characters correctly.
2441    
2442    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2443        character classes. PCRE was not treating the sequence [:...:] as a
2444        character class unless the ... were all letters. Perl, however, seems to
2445        allow any characters between [: and :], though of course it rejects as
2446        unknown any "names" that contain non-letters, because all the known class
2447        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2448        for example, whereas PCRE did not - it did not recognize a POSIX character
2449        class. This seemed a bit dangerous, so the code has been changed to be
2450        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2451        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2452        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2453        Perl does, and where it didn't before.
2454    
2455    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2456        Windows environments %n is disabled by default.
2457    
2458    
2459    Version 7.4 21-Sep-07
2460    ---------------------
2461    
2462    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2463        means that a class such as [\s] counted as "explicit reference to CR or
2464        LF". That isn't really right - the whole point of the change was to try to
2465        help when there was an actual mention of one of the two characters. So now
2466        the change happens only if \r or \n (or a literal CR or LF) character is
2467        encountered.
2468    
2469    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2470        of both had grown to the point where there were only 3 bits left.
2471        Fortunately, there was spare space in the data structure, and so I have
2472        moved the internal flags into a new 16-bit field to free up more option
2473        bits.
2474    
2475    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2476        but did not set the internal JCHANGED flag - either of these is enough to
2477        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2478        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2479        start sets both bits.
2480    
2481    4.  Added options (at build time, compile time, exec time) to change \R from
2482        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2483    
2484    5.  doc/pcresyntax.html was missing from the distribution.
2485    
2486    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2487        compatibility, even though it is no longer used.
2488    
2489    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2490        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2491        windows.h file is present (where different names are used). [This was
2492        reversed later after testing - see 16 below.]
2493    
2494    8.  Changed all #include <config.h> to #include "config.h". There were also
2495        some further <pcre.h> cases that I changed to "pcre.h".
2496    
2497    9.  When pcregrep was used with the --colour option, it missed the line ending
2498        sequence off the lines that it output.
2499    
2500    10. It was pointed out to me that arrays of string pointers cause lots of
2501        relocations when a shared library is dynamically loaded. A technique of
2502        using a single long string with a table of offsets can drastically reduce
2503        these. I have refactored PCRE in four places to do this. The result is
2504        dramatic:
2505    
2506          Originally:                          290
2507          After changing UCP table:            187
2508          After changing error message table:   43
2509          After changing table of "verbs"       36
2510          After changing table of Posix names   22
2511    
2512        Thanks to the folks working on Gregex for glib for this insight.
2513    
2514    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2515        unicode-properties was also set.
2516    
2517    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2518    
2519    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2520        checked only for CRLF.
2521    
2522    14. Added casts to pcretest.c to avoid compiler warnings.
2523    
2524    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2525    
2526    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2527        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2528        entirely. This removes changes made in 7 above.
2529    
2530    17. The CMake files have been updated, and there is now more information about
2531        building with CMake in the NON-UNIX-USE document.
2532    
2533    
2534    Version 7.3 28-Aug-07
2535    ---------------------
2536    
2537     1. In the rejigging of the build system that eventually resulted in 7.1, the
2538        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2539        brackets there is not right, since it causes compilers to look for an
2540        installed pcre.h, not the version that is in the source that is being
2541        compiled (which of course may be different). I have changed it back to:
2542    
2543          #include "pcre.h"
2544    
2545        I have a vague recollection that the change was concerned with compiling in
2546        different directories, but in the new build system, that is taken care of
2547        by the VPATH setting the Makefile.
2548    
2549     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2550        when the subject happened to end in the byte 0x85 (e.g. if the last
2551        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2552        characters but of course it shouldn't be taken as a newline when it is part
2553        of another character. The bug was that, for an unlimited repeat of . in
2554        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2555        characters when looking for a newline.
2556    
2557     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2558    
2559     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2560        in debug output.
2561    
2562     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2563        long printing in the pcrecpp unittest when running under MinGW.
2564    
2565     6. ESC_K was left out of the EBCDIC table.
2566    
2567     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2568        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2569        limit also applies to "virtual nesting" when a pattern is recursive, and in
2570        this case 1000 isn't so big. I have been able to remove this limit at the
2571        expense of backing off one optimization in certain circumstances. Normally,
2572        when pcre_exec() would call its internal match() function recursively and
2573        immediately return the result unconditionally, it uses a "tail recursion"
2574        feature to save stack. However, when a subpattern that can match an empty
2575        string has an unlimited repetition quantifier, it no longer makes this
2576        optimization. That gives it a stack frame in which to save the data for
2577        checking that an empty string has been matched. Previously this was taken
2578        from the 1000-entry workspace that had been reserved. So now there is no
2579        explicit limit, but more stack is used.
2580    
2581     8. Applied Daniel's patches to solve problems with the import/export magic
2582        syntax that is required for Windows, and which was going wrong for the
2583        pcreposix and pcrecpp parts of the library. These were overlooked when this
2584        problem was solved for the main library.
2585    
2586     9. There were some crude static tests to avoid integer overflow when computing
2587        the size of patterns that contain repeated groups with explicit upper
2588        limits. As the maximum quantifier is 65535, the maximum group length was
2589        set at 30,000 so that the product of these two numbers did not overflow a
2590        32-bit integer. However, it turns out that people want to use groups that
2591        are longer than 30,000 bytes (though not repeat them that many times).
2592        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2593        made it possible to implement the integer overflow checks in a much more
2594        dynamic way, which I have now done. The artificial limitation on group
2595        length has been removed - we now have only the limit on the total length of
2596        the compiled pattern, which depends on the LINK_SIZE setting.
2597    
2598    10. Fixed a bug in the documentation for get/copy named substring when
2599        duplicate names are permitted. If none of the named substrings are set, the
2600        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2601        empty string.
2602    
2603    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2604        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2605        because the ] is interpreted as the first data character and the
2606        terminating ] is not found. PCRE has been made compatible with Perl in this
2607        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2608        cause memory overwriting.
2609    
2610    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2611        string has been matched (to stop an infinite loop). It was not recognizing
2612        a conditional subpattern that could match an empty string if that
2613        subpattern was within another subpattern. For example, it looped when
2614        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2615        condition was not nested. This bug has been fixed.
2616    
2617    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2618        past the start of the subject in the presence of bytes with the top bit
2619        set, for example "\x8aBCD".
2620    
2621    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2622        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2623    
2624    14. Optimized (?!) to (*FAIL).
2625    
2626    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2627        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2628        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2629        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2630        does: it's just the validity check that is more restrictive.
2631    
2632    16. Inserted checks for integer overflows during escape sequence (backslash)
2633        processing, and also fixed erroneous offset values for syntax errors during
2634        backslash processing.
2635    
2636    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2637        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2638    
2639    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2640        caused an overrun.
2641    
2642    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2643        something other than just ASCII characters) inside a group that had an
2644        unlimited repeat caused a loop at compile time (while checking to see
2645        whether the group could match an empty string).
2646    
2647    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2648        [\P{Any}] did so. (Error in the code for printing property names.)
2649    
2650    21. An orphan \E inside a character class could cause a crash.
2651    
2652    22. A repeated capturing bracket such as (A)? could cause a wild memory
2653        reference during compilation.
2654    
2655    23. There are several functions in pcre_compile() that scan along a compiled
2656        expression for various reasons (e.g. to see if it's fixed length for look
2657        behind). There were bugs in these functions when a repeated \p or \P was
2658        present in the pattern. These operators have additional parameters compared
2659        with \d, etc, and these were not being taken into account when moving along
2660        the compiled data. Specifically:
2661    
2662        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2663            length.
2664    
2665        (b) An item such as \pL+ within a repeated group could cause crashes or
2666            loops.
2667    
2668        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2669            "reference to non-existent subpattern" error.
2670    
2671        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2672    
2673    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2674        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2675    
2676    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2677    
2678    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2679        character were causing crashes (broken optimization).
2680    
2681    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2682        \p or \P) caused a compile-time loop.
2683    
2684    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2685        break. For example, the unstudied pattern [\r\n]A does not match the string
2686        "\r\nA" because change 7.0/46 below moves the current point on by two
2687        characters after failing to match at the start. However, the pattern \nA
2688        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2689        the same is true. There doesn't seem any very clean way out of this, but
2690        what I have chosen to do makes the common cases work: PCRE now takes note
2691        of whether there can be an explicit match for \r or \n anywhere in the
2692        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2693        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2694        pattern has explicit CR or LF references.
2695    
2696    29. Added (*CR) etc for changing newline setting at start of pattern.
2697    
2698    
2699    Version 7.2 19-Jun-07
2700    ---------------------
2701    
2702     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2703        which is apparently normally available under Windows.
2704    
2705     2. Re-jig the pcregrep tests with different newline settings in an attempt
2706        to make them independent of the local environment's newline setting.
2707    
2708     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2709    
2710     4. Some of the "internals" tests were previously cut out when the link size
2711        was not 2, because the output contained actual offsets. The recent new
2712        "Z" feature of pcretest means that these can be cut out, making the tests
2713        usable with all link sizes.
2714    
2715     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2716        stack recursion. This gives a massive performance boost under BSD, but just
2717        a small improvement under Linux. However, it saves one field in the frame
2718        in all cases.
2719    
2720     6. Added more features from the forthcoming Perl 5.10:
2721    
2722        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2723            recursion call. It refers to the nth most recently opened parentheses.
2724    
2725        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2726            to be opened parentheses.
2727    
2728        (c) Conditions that refer to capturing parentheses can be specified
2729            relatively, for example, (?(-2)... or (?(+3)...
2730    
2731        (d) \K resets the start of the current match so that everything before
2732            is not part of it.
2733    
2734        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2735    
2736        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2737            reference syntax.
2738    
2739        (g) (?| introduces a group in which the numbering of parentheses in each
2740            alternative starts with the same number.
2741    
2742        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2743    
2744     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2745        PCRE_INFO_JCHANGED.
2746    
2747     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2748        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2749        for detecting groups that can match an empty string.
2750    
2751     9. A pattern with a very large number of alternatives (more than several
2752        hundred) was running out of internal workspace during the pre-compile
2753        phase, where pcre_compile() figures out how much memory will be needed. A
2754        bit of new cunning has reduced the workspace needed for groups with
2755        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2756        workspace instead of running out of the 4096 that are available.
2757    
2758    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2759    
2760    11. Applied patch from Google to remove an optimization that didn't quite work.
2761        The report of the bug said:
2762    
2763          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2764          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2765          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2766    
2767    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2768        it matched the wrong number of bytes.
2769    
2770    
2771    Version 7.1 24-Apr-07
2772  ---------------------  ---------------------
2773    
2774   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 2800  Version 7.1 12-Mar-07
2800   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
2801      Weber: (1) pcre_internal.h was missing some function renames; (2) updated      Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2802      makevp.bat for the current PCRE, using the additional files      makevp.bat for the current PCRE, using the additional files
2803      makevp-c.txt, makevp-l.txt, and pcregexp.pas.      makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2804    
2805   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
2806      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 115  Version 7.1 12-Mar-07 Line 2882  Version 7.1 12-Mar-07
2882    
2883  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
2884      segmentation fault instead of returning the appropriate error message.      segmentation fault instead of returning the appropriate error message.
2885    
2886  15. In multiline mode when the newline sequence was set to "any", the pattern  15. In multiline mode when the newline sequence was set to "any", the pattern
2887      ^$ would give a match between the \r and \n of a subject such as "A\r\nB".      ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2888      This doesn't seem right; it now treats the CRLF combination as the line      This doesn't seem right; it now treats the CRLF combination as the line
2889      ending, and so does not match in that case. It's only a pattern such as ^$      ending, and so does not match in that case. It's only a pattern such as ^$
2890      that would hit this one: something like ^ABC$ would have failed after \r      that would hit this one: something like ^ABC$ would have failed after \r
2891      and then tried again after \r\n.      and then tried again after \r\n.
2892    
2893  16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"  16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2894      in an attempt to make files that differ only in their line terminators      in an attempt to make files that differ only in their line terminators
2895      compare equal. This works on Linux.      compare equal. This works on Linux.
2896    
2897    17. Under certain error circumstances pcregrep might try to free random memory
2898        as it exited. This is now fixed, thanks to valgrind.
2899    
2900    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2901        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2902        was because its rules for how to advance for /g after matching an empty
2903        string at the end of a line did not allow for this case. They now check for
2904        it specially.
2905    
2906    20. pcretest is supposed to handle patterns and data of any length, by
2907        extending its buffers when necessary. It was getting this wrong when the
2908        buffer for a data line had to be extended.
2909    
2910    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2911        CRLF as a newline sequence.
2912    
2913    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2914        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2915        I have nevertheless tidied it up.
2916    
2917    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2918    
2919    24. Added a man page for pcre-config.
2920    
2921    
2922  Version 7.0 19-Dec-06  Version 7.0 19-Dec-06

Legend:
Removed from v.131  
changed lines
  Added in v.1464

  ViewVC Help
Powered by ViewVC 1.1.5