ViewVC logotype

Diff of /code/trunk/ChangeLog

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

revision 45 by nigel, Sat Feb 24 21:39:25 2007 UTC revision 1431 by ph10, Thu Jan 2 17:41:28 2014 UTC
# Line 1  Line 1 
1  ChangeLog for PCRE  ChangeLog for PCRE
2  ------------------  ------------------
4    Version 8.35-RC1 xx-xxxx-201x
5    -----------------------------
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.
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.
15    3.  Got rid of some compiler warnings for potentially uninitialized variables
16        that show up only when compiled with -O2.
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.
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.
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.
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.
41    Version 8.34 15-December-2013
42    -----------------------------
44    1.  Add pcre[16|32]_jit_free_unused_memory to forcibly free unused JIT
45        executable memory. Patch inspired by Carsten Klein.
47    2.  ./configure --enable-coverage defined SUPPORT_GCOV in config.h, although
48        this macro is never tested and has no effect, because the work to support
49        coverage involves only compiling and linking options and special targets in
50        the Makefile. The comment in config.h implied that defining the macro would
51        enable coverage support, which is totally false. There was also support for
52        setting this macro in the CMake files (my fault, I just copied it from
53        configure). SUPPORT_GCOV has now been removed.
55    3.  Make a small performance improvement in strlen16() and strlen32() in
56        pcretest.
58    4.  Change 36 for 8.33 left some unreachable statements in pcre_exec.c,
59        detected by the Solaris compiler (gcc doesn't seem to be able to diagnose
60        these cases). There was also one in pcretest.c.
62    5.  Cleaned up a "may be uninitialized" compiler warning in pcre_exec.c.
64    6.  In UTF mode, the code for checking whether a group could match an empty
65        string (which is used for indefinitely repeated groups to allow for
66        breaking an infinite loop) was broken when the group contained a repeated
67        negated single-character class with a character that occupied more than one
68        data item and had a minimum repetition of zero (for example, [^\x{100}]* in
69        UTF-8 mode). The effect was undefined: the group might or might not be
70        deemed as matching an empty string, or the program might have crashed.
72    7.  The code for checking whether a group could match an empty string was not
73        recognizing that \h, \H, \v, \V, and \R must match a character.
75    8.  Implemented PCRE_INFO_MATCH_EMPTY, which yields 1 if the pattern can match
76        an empty string. If it can, pcretest shows this in its information output.
78    9.  Fixed two related bugs that applied to Unicode extended grapheme clusters
79        that were repeated with a maximizing qualifier (e.g. \X* or \X{2,5}) when
80        matched by pcre_exec() without using JIT:
82        (a) If the rest of the pattern did not match after a maximal run of
83            grapheme clusters, the code for backing up to try with fewer of them
84            did not always back up over a full grapheme when characters that do not
85            have the modifier quality were involved, e.g. Hangul syllables.
87        (b) If the match point in a subject started with modifier character, and
88            there was no match, the code could incorrectly back up beyond the match
89            point, and potentially beyond the first character in the subject,
90            leading to a segfault or an incorrect match result.
92    10. A conditional group with an assertion condition could lead to PCRE
93        recording an incorrect first data item for a match if no other first data
94        item was recorded. For example, the pattern (?(?=ab)ab) recorded "a" as a
95        first data item, and therefore matched "ca" after "c" instead of at the
96        start.
98    11. Change 40 for 8.33 (allowing pcregrep to find empty strings) showed up a
99        bug that caused the command "echo a | ./pcregrep -M '|a'" to loop.
101    12. The source of pcregrep now includes z/OS-specific code so that it can be
102        compiled for z/OS as part of the special z/OS distribution.
104    13. Added the -T and -TM options to pcretest.
106    14. The code in pcre_compile.c for creating the table of named capturing groups
107        has been refactored. Instead of creating the table dynamically during the
108        actual compiling pass, the information is remembered during the pre-compile
109        pass (on the stack unless there are more than 20 named groups, in which
110        case malloc() is used) and the whole table is created before the actual
111        compile happens. This has simplified the code (it is now nearly 150 lines
112        shorter) and prepared the way for better handling of references to groups
113        with duplicate names.
115    15. A back reference to a named subpattern when there is more than one of the
116        same name now checks them in the order in which they appear in the pattern.
117        The first one that is set is used for the reference. Previously only the
118        first one was inspected. This change makes PCRE more compatible with Perl.
120    16. Unicode character properties were updated from Unicode 6.3.0.
122    17. The compile-time code for auto-possessification has been refactored, based
123        on a patch by Zoltan Herczeg. It now happens after instead of during
124        compilation. The code is cleaner, and more cases are handled. The option
125        PCRE_NO_AUTO_POSSESS is added for testing purposes, and the -O and /O
126        options in pcretest are provided to set it. It can also be set by
127        (*NO_AUTO_POSSESS) at the start of a pattern.
129    18. The character VT has been added to the default ("C" locale) set of
130        characters that match \s and are generally treated as white space,
131        following this same change in Perl 5.18. There is now no difference between
132        "Perl space" and "POSIX space". Whether VT is treated as white space in
133        other locales depends on the locale.
135    19. The code for checking named groups as conditions, either for being set or
136        for being recursed, has been refactored (this is related to 14 and 15
137        above). Processing unduplicated named groups should now be as fast at
138        numerical groups, and processing duplicated groups should be faster than
139        before.
141    20. Two patches to the CMake build system, by Alexander Barkov:
143          (1) Replace the "source" command by "." in CMakeLists.txt because
144              "source" is a bash-ism.
146          (2) Add missing HAVE_STDINT_H and HAVE_INTTYPES_H to config-cmake.h.in;
147              without these the CMake build does not work on Solaris.
149    21. Perl has changed its handling of \8 and \9. If there is no previously
150        encountered capturing group of those numbers, they are treated as the
151        literal characters 8 and 9 instead of a binary zero followed by the
152        literals. PCRE now does the same.
154    22. Following Perl, added \o{} to specify codepoints in octal, making it
155        possible to specify values greater than 0777 and also making them
156        unambiguous.
158    23. Perl now gives an error for missing closing braces after \x{... instead of
159        treating the string as literal. PCRE now does the same.
161    24. RunTest used to grumble if an inappropriate test was selected explicitly,
162        but just skip it when running all tests. This make it awkward to run ranges
163        of tests when one of them was inappropriate. Now it just skips any
164        inappropriate tests, as it always did when running all tests.
166    25. If PCRE_AUTO_CALLOUT and PCRE_UCP were set for a pattern that contained
167        character types such as \d or \w, too many callouts were inserted, and the
168        data that they returned was rubbish.
170    26. In UCP mode, \s was not matching two of the characters that Perl matches,
171        namely NEL (U+0085) and MONGOLIAN VOWEL SEPARATOR (U+180E), though they
172        were matched by \h. The code has now been refactored so that the lists of
173        the horizontal and vertical whitespace characters used for \h and \v (which
174        are defined only in one place) are now also used for \s.
176    27. Add JIT support for the 64 bit TileGX architecture.
177        Patch by Jiong Wang (Tilera Corporation).
179    28. Possessive quantifiers for classes (both explicit and automatically
180        generated) now use special opcodes instead of wrapping in ONCE brackets.
182    29. Whereas an item such as A{4}+ ignored the possessivenes of the quantifier
183        (because it's meaningless), this was not happening when PCRE_CASELESS was
184        set. Not wrong, but inefficient.
186    30. Updated perltest.pl to add /u (force Unicode mode) when /W (use Unicode
187        properties for \w, \d, etc) is present in a test regex. Otherwise if the
188        test contains no characters greater than 255, Perl doesn't realise it
189        should be using Unicode semantics.
191    31. Upgraded the handling of the POSIX classes [:graph:], [:print:], and
192        [:punct:] when PCRE_UCP is set so as to include the same characters as Perl
193        does in Unicode mode.
195    32. Added the "forbid" facility to pcretest so that putting tests into the
196        wrong test files can sometimes be quickly detected.
198    33. There is now a limit (default 250) on the depth of nesting of parentheses.
199        This limit is imposed to control the amount of system stack used at compile
200        time. It can be changed at build time by --with-parens-nest-limit=xxx or
201        the equivalent in CMake.
203    34. Character classes such as [A-\d] or [a-[:digit:]] now cause compile-time
204        errors. Perl warns for these when in warning mode, but PCRE has no facility
205        for giving warnings.
207    35. Change 34 for 8.13 allowed quantifiers on assertions, because Perl does.
208        However, this was not working for (?!) because it is optimized to (*FAIL),
209        for which PCRE does not allow quantifiers. The optimization is now disabled
210        when a quantifier follows (?!). I can't see any use for this, but it makes
211        things uniform.
213    36. Perl no longer allows group names to start with digits, so I have made this
214        change also in PCRE. It simplifies the code a bit.
216    37. In extended mode, Perl ignores spaces before a + that indicates a
217        possessive quantifier. PCRE allowed a space before the quantifier, but not
218        before the possessive +. It now does.
220    38. The use of \K (reset reported match start) within a repeated possessive
221        group such as (a\Kb)*+ was not working.
223    40. Document that the same character tables must be used at compile time and
224        run time, and that the facility to pass tables to pcre_exec() and
225        pcre_dfa_exec() is for use only with saved/restored patterns.
227    41. Applied Jeff Trawick's patch CMakeLists.txt, which "provides two new
228        features for Builds with MSVC:
230        1. Support pcre.rc and/or pcreposix.rc (as is already done for MinGW
231           builds). The .rc files can be used to set FileDescription and many other
232           attributes.
234        2. Add an option (-DINSTALL_MSVC_PDB) to enable installation of .pdb files.
235           This allows higher-level build scripts which want .pdb files to avoid
236           hard-coding the exact files needed."
238    42. Added support for [[:<:]] and [[:>:]] as used in the BSD POSIX library to
239        mean "start of word" and "end of word", respectively, as a transition aid.
241    43. A minimizing repeat of a class containing codepoints greater than 255 in
242        non-UTF 16-bit or 32-bit modes caused an internal error when PCRE was
243        compiled to use the heap for recursion.
245    44. Got rid of some compiler warnings for unused variables when UTF but not UCP
246        is configured.
249    Version 8.33 28-May-2013
250    ------------------------
252    1.  Added 'U' to some constants that are compared to unsigned integers, to
253        avoid compiler signed/unsigned warnings. Added (int) casts to unsigned
254        variables that are added to signed variables, to ensure the result is
255        signed and can be negated.
257    2.  Applied patch by Daniel Richard G for quashing MSVC warnings to the
258        CMake config files.
260    3.  Revise the creation of config.h.generic so that all boolean macros are
261        #undefined, whereas non-boolean macros are #ifndef/#endif-ed. This makes
262        overriding via -D on the command line possible.
264    4.  Changing the definition of the variable "op" in pcre_exec.c from pcre_uchar
265        to unsigned int is reported to make a quite noticeable speed difference in
266        a specific Windows environment. Testing on Linux did also appear to show
267        some benefit (and it is clearly not harmful). Also fixed the definition of
268        Xop which should be unsigned.
270    5.  Related to (4), changing the definition of the intermediate variable cc
271        in repeated character loops from pcre_uchar to pcre_uint32 also gave speed
272        improvements.
274    6.  Fix forward search in JIT when link size is 3 or greater. Also removed some
275        unnecessary spaces.
277    7.  Adjust autogen.sh and configure.ac to lose warnings given by automake 1.12
278        and later.
280    8.  Fix two buffer over read issues in 16 and 32 bit modes. Affects JIT only.
282    9.  Optimizing fast_forward_start_bits in JIT.
284    10. Adding support for callouts in JIT, and fixing some issues revealed
285        during this work. Namely:
287        (a) Unoptimized capturing brackets incorrectly reset on backtrack.
289        (b) Minimum length was not checked before the matching is started.
291    11. The value of capture_last that is passed to callouts was incorrect in some
292        cases when there was a capture on one path that was subsequently abandoned
293        after a backtrack. Also, the capture_last value is now reset after a
294        recursion, since all captures are also reset in this case.
296    12. The interpreter no longer returns the "too many substrings" error in the
297        case when an overflowing capture is in a branch that is subsequently
298        abandoned after a backtrack.
300    13. In the pathological case when an offset vector of size 2 is used, pcretest
301        now prints out the matched string after a yield of 0 or 1.
303    14. Inlining subpatterns in recursions, when certain conditions are fulfilled.
304        Only supported by the JIT compiler at the moment.
306    15. JIT compiler now supports 32 bit Macs thanks to Lawrence Velazquez.
308    16. Partial matches now set offsets[2] to the "bumpalong" value, that is, the
309        offset of the starting point of the matching process, provided the offsets
310        vector is large enough.
312    17. The \A escape now records a lookbehind value of 1, though its execution
313        does not actually inspect the previous character. This is to ensure that,
314        in partial multi-segment matching, at least one character from the old
315        segment is retained when a new segment is processed. Otherwise, if there
316        are no lookbehinds in the pattern, \A might match incorrectly at the start
317        of a new segment.
319    18. Added some #ifdef __VMS code into pcretest.c to help VMS implementations.
321    19. Redefined some pcre_uchar variables in pcre_exec.c as pcre_uint32; this
322        gives some modest performance improvement in 8-bit mode.
324    20. Added the PCRE-specific property \p{Xuc} for matching characters that can
325        be expressed in certain programming languages using Universal Character
326        Names.
328    21. Unicode validation has been updated in the light of Unicode Corrigendum #9,
329        which points out that "non characters" are not "characters that may not
330        appear in Unicode strings" but rather "characters that are reserved for
331        internal use and have only local meaning".
333    22. When a pattern was compiled with automatic callouts (PCRE_AUTO_CALLOUT) and
334        there was a conditional group that depended on an assertion, if the
335        assertion was false, the callout that immediately followed the alternation
336        in the condition was skipped when pcre_exec() was used for matching.
338    23. Allow an explicit callout to be inserted before an assertion that is the
339        condition for a conditional group, for compatibility with automatic
340        callouts, which always insert a callout at this point.
342    24. In 8.31, (*COMMIT) was confined to within a recursive subpattern. Perl also
343        confines (*SKIP) and (*PRUNE) in the same way, and this has now been done.
345    25. (*PRUNE) is now supported by the JIT compiler.
347    26. Fix infinite loop when /(?<=(*SKIP)ac)a/ is matched against aa.
349    27. Fix the case where there are two or more SKIPs with arguments that may be
350        ignored.
352    28. (*SKIP) is now supported by the JIT compiler.
354    29. (*THEN) is now supported by the JIT compiler.
356    30. Update RunTest with additional test selector options.
358    31. The way PCRE handles backtracking verbs has been changed in two ways.
360        (1) Previously, in something like (*COMMIT)(*SKIP), COMMIT would override
361        SKIP. Now, PCRE acts on whichever backtracking verb is reached first by
362        backtracking. In some cases this makes it more Perl-compatible, but Perl's
363        rather obscure rules do not always do the same thing.
365        (2) Previously, backtracking verbs were confined within assertions. This is
366        no longer the case for positive assertions, except for (*ACCEPT). Again,
367        this sometimes improves Perl compatibility, and sometimes does not.
369    32. A number of tests that were in test 2 because Perl did things differently
370        have been moved to test 1, because either Perl or PCRE has changed, and
371        these tests are now compatible.
373    32. Backtracking control verbs are now handled in the same way in JIT and
374        interpreter.
376    33. An opening parenthesis in a MARK/PRUNE/SKIP/THEN name in a pattern that
377        contained a forward subroutine reference caused a compile error.
379    34. Auto-detect and optimize limited repetitions in JIT.
381    35. Implement PCRE_NEVER_UTF to lock out the use of UTF, in particular,
382        blocking (*UTF) etc.
384    36. In the interpreter, maximizing pattern repetitions for characters and
385        character types now use tail recursion, which reduces stack usage.
387    37. The value of the max lookbehind was not correctly preserved if a compiled
388        and saved regex was reloaded on a host of different endianness.
390    38. Implemented (*LIMIT_MATCH) and (*LIMIT_RECURSION). As part of the extension
391        of the compiled pattern block, expand the flags field from 16 to 32 bits
392        because it was almost full.
394    39. Try madvise first before posix_madvise.
396    40. Change 7 for PCRE 7.9 made it impossible for pcregrep to find empty lines
397        with a pattern such as ^$. It has taken 4 years for anybody to notice! The
398        original change locked out all matches of empty strings. This has been
399        changed so that one match of an empty string per line is recognized.
400        Subsequent searches on the same line (for colouring or for --only-matching,
401        for example) do not recognize empty strings.
403    41. Applied a user patch to fix a number of spelling mistakes in comments.
405    42. Data lines longer than 65536 caused pcretest to crash.
407    43. Clarified the data type for length and startoffset arguments for pcre_exec
408        and pcre_dfa_exec in the function-specific man pages, where they were
409        explicitly stated to be in bytes, never having been updated. I also added
410        some clarification to the pcreapi man page.
412    44. A call to pcre_dfa_exec() with an output vector size less than 2 caused
413        a segmentation fault.
416    Version 8.32 30-November-2012
417    -----------------------------
419    1.  Improved JIT compiler optimizations for first character search and single
420        character iterators.
422    2.  Supporting IBM XL C compilers for PPC architectures in the JIT compiler.
423        Patch by Daniel Richard G.
425    3.  Single character iterator optimizations in the JIT compiler.
427    4.  Improved JIT compiler optimizations for character ranges.
429    5.  Rename the "leave" variable names to "quit" to improve WinCE compatibility.
430        Reported by Giuseppe D'Angelo.
432    6.  The PCRE_STARTLINE bit, indicating that a match can occur only at the start
433        of a line, was being set incorrectly in cases where .* appeared inside
434        atomic brackets at the start of a pattern, or where there was a subsequent
435        *PRUNE or *SKIP.
437    7.  Improved instruction cache flush for POWER/PowerPC.
438        Patch by Daniel Richard G.
440    8.  Fixed a number of issues in pcregrep, making it more compatible with GNU
441        grep:
443        (a) There is now no limit to the number of patterns to be matched.
445        (b) An error is given if a pattern is too long.
447        (c) Multiple uses of --exclude, --exclude-dir, --include, and --include-dir
448            are now supported.
450        (d) --exclude-from and --include-from (multiple use) have been added.
452        (e) Exclusions and inclusions now apply to all files and directories, not
453            just to those obtained from scanning a directory recursively.
455        (f) Multiple uses of -f and --file-list are now supported.
457        (g) In a Windows environment, the default for -d has been changed from
458            "read" (the GNU grep default) to "skip", because otherwise the presence
459            of a directory in the file list provokes an error.
461        (h) The documentation has been revised and clarified in places.
463    9.  Improve the matching speed of capturing brackets.
465    10. Changed the meaning of \X so that it now matches a Unicode extended
466        grapheme cluster.
468    11. Patch by Daniel Richard G to the autoconf files to add a macro for sorting
469        out POSIX threads when JIT support is configured.
471    12. Added support for PCRE_STUDY_EXTRA_NEEDED.
473    13. In the POSIX wrapper regcomp() function, setting re_nsub field in the preg
474        structure could go wrong in environments where size_t is not the same size
475        as int.
477    14. Applied user-supplied patch to pcrecpp.cc to allow PCRE_NO_UTF8_CHECK to be
478        set.
480    15. The EBCDIC support had decayed; later updates to the code had included
481        explicit references to (e.g.) \x0a instead of CHAR_LF. There has been a
482        general tidy up of EBCDIC-related issues, and the documentation was also
483        not quite right. There is now a test that can be run on ASCII systems to
484        check some of the EBCDIC-related things (but is it not a full test).
486    16. The new PCRE_STUDY_EXTRA_NEEDED option is now used by pcregrep, resulting
487        in a small tidy to the code.
489    17. Fix JIT tests when UTF is disabled and both 8 and 16 bit mode are enabled.
491    18. If the --only-matching (-o) option in pcregrep is specified multiple
492        times, each one causes appropriate output. For example, -o1 -o2 outputs the
493        substrings matched by the 1st and 2nd capturing parentheses. A separating
494        string can be specified by --om-separator (default empty).
496    19. Improving the first n character searches.
498    20. Turn case lists for horizontal and vertical white space into macros so that
499        they are defined only once.
501    21. This set of changes together give more compatible Unicode case-folding
502        behaviour for characters that have more than one other case when UCP
503        support is available.
505        (a) The Unicode property table now has offsets into a new table of sets of
506            three or more characters that are case-equivalent. The MultiStage2.py
507            script that generates these tables (the pcre_ucd.c file) now scans
508            CaseFolding.txt instead of UnicodeData.txt for character case
509            information.
511        (b) The code for adding characters or ranges of characters to a character
512            class has been abstracted into a generalized function that also handles
513            case-independence. In UTF-mode with UCP support, this uses the new data
514            to handle characters with more than one other case.
516        (c) A bug that is fixed as a result of (b) is that codepoints less than 256
517            whose other case is greater than 256 are now correctly matched
518            caselessly. Previously, the high codepoint matched the low one, but not
519            vice versa.
521        (d) The processing of \h, \H, \v, and \ in character classes now makes use
522            of the new class addition function, using character lists defined as
523            macros alongside the case definitions of 20 above.
525        (e) Caseless back references now work with characters that have more than
526            one other case.
528        (f) General caseless matching of characters with more than one other case
529            is supported.
531    22. Unicode character properties were updated from Unicode 6.2.0
533    23. Improved CMake support under Windows. Patch by Daniel Richard G.
535    24. Add support for 32-bit character strings, and UTF-32
537    25. Major JIT compiler update (code refactoring and bugfixing).
538        Experimental Sparc 32 support is added.
540    26. Applied a modified version of Daniel Richard G's patch to create
541        pcre.h.generic and config.h.generic by "make" instead of in the
542        PrepareRelease script.
544    27. Added a definition for CHAR_NULL (helpful for the z/OS port), and use it in
545        pcre_compile.c when checking for a zero character.
547    28. Introducing a native interface for JIT. Through this interface, the compiled
548        machine code can be directly executed. The purpose of this interface is to
549        provide fast pattern matching, so several sanity checks are not performed.
550        However, feature tests are still performed. The new interface provides
551        1.4x speedup compared to the old one.
553    29. If pcre_exec() or pcre_dfa_exec() was called with a negative value for
554        the subject string length, the error given was PCRE_ERROR_BADOFFSET, which
555        was confusing. There is now a new error PCRE_ERROR_BADLENGTH for this case.
557    30. In 8-bit UTF-8 mode, pcretest failed to give an error for data codepoints
558        greater than 0x7fffffff (which cannot be represented in UTF-8, even under
559        the "old" RFC 2279). Instead, it ended up passing a negative length to
560        pcre_exec().
562    31. Add support for GCC's visibility feature to hide internal functions.
564    32. Running "pcretest -C pcre8" or "pcretest -C pcre16" gave a spurious error
565        "unknown -C option" after outputting 0 or 1.
567    33. There is now support for generating a code coverage report for the test
568        suite in environments where gcc is the compiler and lcov is installed. This
569        is mainly for the benefit of the developers.
571    34. If PCRE is built with --enable-valgrind, certain memory regions are marked
572        unaddressable using valgrind annotations, allowing valgrind to detect
573        invalid memory accesses. This is mainly for the benefit of the developers.
575    25. (*UTF) can now be used to start a pattern in any of the three libraries.
577    26. Give configure error if --enable-cpp but no C++ compiler found.
580    Version 8.31 06-July-2012
581    -------------------------
583    1.  Fixing a wrong JIT test case and some compiler warnings.
585    2.  Removed a bashism from the RunTest script.
587    3.  Add a cast to pcre_exec.c to fix the warning "unary minus operator applied
588        to unsigned type, result still unsigned" that was given by an MS compiler
589        on encountering the code "-sizeof(xxx)".
591    4.  Partial matching support is added to the JIT compiler.
593    5.  Fixed several bugs concerned with partial matching of items that consist
594        of more than one character:
596        (a) /^(..)\1/ did not partially match "aba" because checking references was
597            done on an "all or nothing" basis. This also applied to repeated
598            references.
600        (b) \R did not give a hard partial match if \r was found at the end of the
601            subject.
603        (c) \X did not give a hard partial match after matching one or more
604            characters at the end of the subject.
606        (d) When newline was set to CRLF, a pattern such as /a$/ did not recognize
607            a partial match for the string "\r".
609        (e) When newline was set to CRLF, the metacharacter "." did not recognize
610            a partial match for a CR character at the end of the subject string.
612    6.  If JIT is requested using /S++ or -s++ (instead of just /S+ or -s+) when
613        running pcretest, the text "(JIT)" added to the output whenever JIT is
614        actually used to run the match.
616    7.  Individual JIT compile options can be set in pcretest by following -s+[+]
617        or /S+[+] with a digit between 1 and 7.
619    8.  OP_NOT now supports any UTF character not just single-byte ones.
621    9.  (*MARK) control verb is now supported by the JIT compiler.
623    10. The command "./RunTest list" lists the available tests without actually
624        running any of them. (Because I keep forgetting what they all are.)
628    12. Applied a (slightly modified) user-supplied patch that improves performance
629        when the heap is used for recursion (compiled with --disable-stack-for-
630        recursion). Instead of malloc and free for each heap frame each time a
631        logical recursion happens, frames are retained on a chain and re-used where
632        possible. This sometimes gives as much as 30% improvement.
634    13. As documented, (*COMMIT) is now confined to within a recursive subpattern
635        call.
637    14. As documented, (*COMMIT) is now confined to within a positive assertion.
639    15. It is now possible to link pcretest with libedit as an alternative to
640        libreadline.
642    16. (*COMMIT) control verb is now supported by the JIT compiler.
644    17. The Unicode data tables have been updated to Unicode 6.1.0.
646    18. Added --file-list option to pcregrep.
648    19. Added binary file support to pcregrep, including the -a, --binary-files,
649        -I, and --text options.
651    20. The madvise function is renamed for posix_madvise for QNX compatibility
652        reasons. Fixed by Giuseppe D'Angelo.
654    21. Fixed a bug for backward assertions with REVERSE 0 in the JIT compiler.
656    22. Changed the option for creating symbolic links for 16-bit man pages from
657        -s to -sf so that re-installing does not cause issues.
659    23. Support PCRE_NO_START_OPTIMIZE in JIT as (*MARK) support requires it.
661    24. Fixed a very old bug in pcretest that caused errors with restarted DFA
662        matches in certain environments (the workspace was not being correctly
663        retained). Also added to pcre_dfa_exec() a simple plausibility check on
664        some of the workspace data at the beginning of a restart.
666    25. \s*\R was auto-possessifying the \s* when it should not, whereas \S*\R
667        was not doing so when it should - probably a typo introduced by SVN 528
668        (change 8.10/14).
670    26. When PCRE_UCP was not set, \w+\x{c4} was incorrectly auto-possessifying the
671        \w+ when the character tables indicated that \x{c4} was a word character.
672        There were several related cases, all because the tests for doing a table
673        lookup were testing for characters less than 127 instead of 255.
675    27. If a pattern contains capturing parentheses that are not used in a match,
676        their slots in the ovector are set to -1. For those that are higher than
677        any matched groups, this happens at the end of processing. In the case when
678        there were back references that the ovector was too small to contain
679        (causing temporary malloc'd memory to be used during matching), and the
680        highest capturing number was not used, memory off the end of the ovector
681        was incorrectly being set to -1. (It was using the size of the temporary
682        memory instead of the true size.)
684    28. To catch bugs like 27 using valgrind, when pcretest is asked to specify an
685        ovector size, it uses memory at the end of the block that it has got.
687    29. Check for an overlong MARK name and give an error at compile time. The
688        limit is 255 for the 8-bit library and 65535 for the 16-bit library.
690    30. JIT compiler update.
692    31. JIT is now supported on jailbroken iOS devices. Thanks for Ruiger
693        Rill for the patch.
695    32. Put spaces around SLJIT_PRINT_D in the JIT compiler. Required by CXX11.
697    33. Variable renamings in the PCRE-JIT compiler. No functionality change.
699    34. Fixed typos in pcregrep: in two places there was SUPPORT_LIBZ2 instead of
700        SUPPORT_LIBBZ2. This caused a build problem when bzip2 but not gzip (zlib)
701        was enabled.
703    35. Improve JIT code generation for greedy plus quantifier.
705    36. When /((?:a?)*)*c/ or /((?>a?)*)*c/ was matched against "aac", it set group
706        1 to "aa" instead of to an empty string. The bug affected repeated groups
707        that could potentially match an empty string.
709    37. Optimizing single character iterators in JIT.
711    38. Wide characters specified with \uxxxx in JavaScript mode are now subject to
712        the same checks as \x{...} characters in non-JavaScript mode. Specifically,
713        codepoints that are too big for the mode are faulted, and in a UTF mode,
714        disallowed codepoints are also faulted.
716    39. If PCRE was compiled with UTF support, in three places in the DFA
717        matcher there was code that should only have been obeyed in UTF mode, but
718        was being obeyed unconditionally. In 8-bit mode this could cause incorrect
719        processing when bytes with values greater than 127 were present. In 16-bit
720        mode the bug would be provoked by values in the range 0xfc00 to 0xdc00. In
721        both cases the values are those that cannot be the first data item in a UTF
722        character. The three items that might have provoked this were recursions,
723        possessively repeated groups, and atomic groups.
725    40. Ensure that libpcre is explicitly listed in the link commands for pcretest
726        and pcregrep, because some OS require shared objects to be explicitly
727        passed to ld, causing the link step to fail if they are not.
729    41. There were two incorrect #ifdefs in pcre_study.c, meaning that, in 16-bit
730        mode, patterns that started with \h* or \R* might be incorrectly matched.
733    Version 8.30 04-February-2012
734    -----------------------------
736    1.  Renamed "isnumber" as "is_a_number" because in some Mac environments this
737        name is defined in ctype.h.
739    2.  Fixed a bug in fixed-length calculation for lookbehinds that would show up
740        only in quite long subpatterns.
742    3.  Removed the function pcre_info(), which has been obsolete and deprecated
743        since it was replaced by pcre_fullinfo() in February 2000.
745    4.  For a non-anchored pattern, if (*SKIP) was given with a name that did not
746        match a (*MARK), and the match failed at the start of the subject, a
747        reference to memory before the start of the subject could occur. This bug
748        was introduced by fix 17 of release 8.21.
750    5.  A reference to an unset group with zero minimum repetition was giving
751        totally wrong answers (in non-JavaScript-compatibility mode). For example,
752        /(another)?(\1?)test/ matched against "hello world test". This bug was
753        introduced in release 8.13.
755    6.  Add support for 16-bit character strings (a large amount of work involving
756        many changes and refactorings).
758    7.  RunGrepTest failed on msys because \r\n was replaced by whitespace when the
759        command "pattern=`printf 'xxx\r\njkl'`" was run. The pattern is now taken
760        from a file.
762    8.  Ovector size of 2 is also supported by JIT based pcre_exec (the ovector size
763        rounding is not applied in this particular case).
765    9.  The invalid Unicode surrogate codepoints U+D800 to U+DFFF are now rejected
766        if they appear, or are escaped, in patterns.
768    10. Get rid of a number of -Wunused-but-set-variable warnings.
770    11. The pattern /(?=(*:x))(q|)/ matches an empty string, and returns the mark
771        "x". The similar pattern /(?=(*:x))((*:y)q|)/ did not return a mark at all.
772        Oddly, Perl behaves the same way. PCRE has been fixed so that this pattern
773        also returns the mark "x". This bug applied to capturing parentheses,
774        non-capturing parentheses, and atomic parentheses. It also applied to some
775        assertions.
777    12. Stephen Kelly's patch to CMakeLists.txt allows it to parse the version
778        information out of configure.ac instead of relying on pcre.h.generic, which
779        is not stored in the repository.
781    13. Applied Dmitry V. Levin's patch for a more portable method for linking with
782        -lreadline.
784    14. ZH added PCRE_CONFIG_JITTARGET; added its output to pcretest -C.
786    15. Applied Graycode's patch to put the top-level frame on the stack rather
787        than the heap when not using the stack for recursion. This gives a
788        performance improvement in many cases when recursion is not deep.
790    16. Experimental code added to "pcretest -C" to output the stack frame size.
793    Version 8.21 12-Dec-2011
794    ------------------------
796    1.  Updating the JIT compiler.
798    2.  JIT compiler now supports OP_NCREF, OP_RREF and OP_NRREF. New test cases
799        are added as well.
801    3.  Fix cache-flush issue on PowerPC (It is still an experimental JIT port).
802        PCRE_EXTRA_TABLES is not suported by JIT, and should be checked before
803        calling _pcre_jit_exec. Some extra comments are added.
805    4.  (*MARK) settings inside atomic groups that do not contain any capturing
806        parentheses, for example, (?>a(*:m)), were not being passed out. This bug
807        was introduced by change 18 for 8.20.
809    5.  Supporting of \x, \U and \u in JavaScript compatibility mode based on the
810        ECMA-262 standard.
812    6.  Lookbehinds such as (?<=a{2}b) that contained a fixed repetition were
813        erroneously being rejected as "not fixed length" if PCRE_CASELESS was set.
814        This bug was probably introduced by change 9 of 8.13.
816    7.  While fixing 6 above, I noticed that a number of other items were being
817        incorrectly rejected as "not fixed length". This arose partly because newer
818        opcodes had not been added to the fixed-length checking code. I have (a)
819        corrected the bug and added tests for these items, and (b) arranged for an
820        error to occur if an unknown opcode is encountered while checking for fixed
821        length instead of just assuming "not fixed length". The items that were
822        rejected were: (*ACCEPT), (*COMMIT), (*FAIL), (*MARK), (*PRUNE), (*SKIP),
823        (*THEN), \h, \H, \v, \V, and single character negative classes with fixed
824        repetitions, e.g. [^a]{3}, with and without PCRE_CASELESS.
826    8.  A possessively repeated conditional subpattern such as (?(?=c)c|d)++ was
827        being incorrectly compiled and would have given unpredicatble results.
829    9.  A possessively repeated subpattern with minimum repeat count greater than
830        one behaved incorrectly. For example, (A){2,}+ behaved as if it was
831        (A)(A)++ which meant that, after a subsequent mismatch, backtracking into
832        the first (A) could occur when it should not.
834    10. Add a cast and remove a redundant test from the code.
836    11. JIT should use pcre_malloc/pcre_free for allocation.
838    12. Updated pcre-config so that it no longer shows -L/usr/lib, which seems
839        best practice nowadays, and helps with cross-compiling. (If the exec_prefix
840        is anything other than /usr, -L is still shown).
842    13. In non-UTF-8 mode, \C is now supported in lookbehinds and DFA matching.
844    14. Perl does not support \N without a following name in a [] class; PCRE now
845        also gives an error.
847    15. If a forward reference was repeated with an upper limit of around 2000,
848        it caused the error "internal error: overran compiling workspace". The
849        maximum number of forward references (including repeats) was limited by the
850        internal workspace, and dependent on the LINK_SIZE. The code has been
851        rewritten so that the workspace expands (via pcre_malloc) if necessary, and
852        the default depends on LINK_SIZE. There is a new upper limit (for safety)
853        of around 200,000 forward references. While doing this, I also speeded up
854        the filling in of repeated forward references.
856    16. A repeated forward reference in a pattern such as (a)(?2){2}(.) was
857        incorrectly expecting the subject to contain another "a" after the start.
859    17. When (*SKIP:name) is activated without a corresponding (*MARK:name) earlier
860        in the match, the SKIP should be ignored. This was not happening; instead
861        the SKIP was being treated as NOMATCH. For patterns such as
862        /A(*MARK:A)A+(*SKIP:B)Z|AAC/ this meant that the AAC branch was never
863        tested.
865    18. The behaviour of (*MARK), (*PRUNE), and (*THEN) has been reworked and is
866        now much more compatible with Perl, in particular in cases where the result
867        is a non-match for a non-anchored pattern. For example, if
868        /b(*:m)f|a(*:n)w/ is matched against "abc", the non-match returns the name
869        "m", where previously it did not return a name. A side effect of this
870        change is that for partial matches, the last encountered mark name is
871        returned, as for non matches. A number of tests that were previously not
872        Perl-compatible have been moved into the Perl-compatible test files. The
873        refactoring has had the pleasing side effect of removing one argument from
874        the match() function, thus reducing its stack requirements.
876    19. If the /S+ option was used in pcretest to study a pattern using JIT,
877        subsequent uses of /S (without +) incorrectly behaved like /S+.
879    21. Retrieve executable code size support for the JIT compiler and fixing
880        some warnings.
882    22. A caseless match of a UTF-8 character whose other case uses fewer bytes did
883        not work when the shorter character appeared right at the end of the
884        subject string.
886    23. Added some (int) casts to non-JIT modules to reduce warnings on 64-bit
887        systems.
889    24. Added PCRE_INFO_JITSIZE to pass on the value from (21) above, and also
890        output it when the /M option is used in pcretest.
892    25. The CheckMan script was not being included in the distribution. Also, added
893        an explicit "perl" to run Perl scripts from the PrepareRelease script
894        because this is reportedly needed in Windows.
896    26. If study data was being save in a file and studying had not found a set of
897        "starts with" bytes for the pattern, the data written to the file (though
898        never used) was taken from uninitialized memory and so caused valgrind to
899        complain.
901    27. Updated RunTest.bat as provided by Sheri Pierce.
903    28. Fixed a possible uninitialized memory bug in pcre_jit_compile.c.
905    29. Computation of memory usage for the table of capturing group names was
906        giving an unnecessarily large value.
909    Version 8.20 21-Oct-2011
910    ------------------------
912    1.  Change 37 of 8.13 broke patterns like [:a]...[b:] because it thought it had
913        a POSIX class. After further experiments with Perl, which convinced me that
914        Perl has bugs and confusions, a closing square bracket is no longer allowed
915        in a POSIX name. This bug also affected patterns with classes that started
916        with full stops.
918    2.  If a pattern such as /(a)b|ac/ is matched against "ac", there is no
919        captured substring, but while checking the failing first alternative,
920        substring 1 is temporarily captured. If the output vector supplied to
921        pcre_exec() was not big enough for this capture, the yield of the function
922        was still zero ("insufficient space for captured substrings"). This cannot
923        be totally fixed without adding another stack variable, which seems a lot
924        of expense for a edge case. However, I have improved the situation in cases
925        such as /(a)(b)x|abc/ matched against "abc", where the return code
926        indicates that fewer than the maximum number of slots in the ovector have
927        been set.
929    3.  Related to (2) above: when there are more back references in a pattern than
930        slots in the output vector, pcre_exec() uses temporary memory during
931        matching, and copies in the captures as far as possible afterwards. It was
932        using the entire output vector, but this conflicts with the specification
933        that only 2/3 is used for passing back captured substrings. Now it uses
934        only the first 2/3, for compatibility. This is, of course, another edge
935        case.
937    4.  Zoltan Herczeg's just-in-time compiler support has been integrated into the
938        main code base, and can be used by building with --enable-jit. When this is
939        done, pcregrep automatically uses it unless --disable-pcregrep-jit or the
940        runtime --no-jit option is given.
942    5.  When the number of matches in a pcre_dfa_exec() run exactly filled the
943        ovector, the return from the function was zero, implying that there were
944        other matches that did not fit. The correct "exactly full" value is now
945        returned.
947    6.  If a subpattern that was called recursively or as a subroutine contained
948        (*PRUNE) or any other control that caused it to give a non-standard return,
949        invalid errors such as "Error -26 (nested recursion at the same subject
950        position)" or even infinite loops could occur.
952    7.  If a pattern such as /a(*SKIP)c|b(*ACCEPT)|/ was studied, it stopped
953        computing the minimum length on reaching *ACCEPT, and so ended up with the
954        wrong value of 1 rather than 0. Further investigation indicates that
955        computing a minimum subject length in the presence of *ACCEPT is difficult
956        (think back references, subroutine calls), and so I have changed the code
957        so that no minimum is registered for a pattern that contains *ACCEPT.
959    8.  If (*THEN) was present in the first (true) branch of a conditional group,
960        it was not handled as intended. [But see 16 below.]
962    9.  Replaced RunTest.bat and CMakeLists.txt with improved versions provided by
963        Sheri Pierce.
965    10. A pathological pattern such as /(*ACCEPT)a/ was miscompiled, thinking that
966        the first byte in a match must be "a".
968    11. Change 17 for 8.13 increased the recursion depth for patterns like
969        /a(?:.)*?a/ drastically. I've improved things by remembering whether a
970        pattern contains any instances of (*THEN). If it does not, the old
971        optimizations are restored. It would be nice to do this on a per-group
972        basis, but at the moment that is not feasible.
974    12. In some environments, the output of pcretest -C is CRLF terminated. This
975        broke RunTest's code that checks for the link size. A single white space
976        character after the value is now allowed for.
978    13. RunTest now checks for the "fr" locale as well as for "fr_FR" and "french".
979        For "fr", it uses the Windows-specific input and output files.
981    14. If (*THEN) appeared in a group that was called recursively or as a
982        subroutine, it did not work as intended. [But see next item.]
984    15. Consider the pattern /A (B(*THEN)C) | D/ where A, B, C, and D are complex
985        pattern fragments (but not containing any | characters). If A and B are
986        matched, but there is a failure in C so that it backtracks to (*THEN), PCRE
987        was behaving differently to Perl. PCRE backtracked into A, but Perl goes to
988        D. In other words, Perl considers parentheses that do not contain any |
989        characters to be part of a surrounding alternative, whereas PCRE was
990        treading (B(*THEN)C) the same as (B(*THEN)C|(*FAIL)) -- which Perl handles
991        differently. PCRE now behaves in the same way as Perl, except in the case
992        of subroutine/recursion calls such as (?1) which have in any case always
993        been different (but PCRE had them first :-).
995    16. Related to 15 above: Perl does not treat the | in a conditional group as
996        creating alternatives. Such a group is treated in the same way as an
997        ordinary group without any | characters when processing (*THEN). PCRE has
998        been changed to match Perl's behaviour.
1000    17. If a user had set PCREGREP_COLO(U)R to something other than 1:31, the
1001        RunGrepTest script failed.
1003    18. Change 22 for version 13 caused atomic groups to use more stack. This is
1004        inevitable for groups that contain captures, but it can lead to a lot of
1005        stack use in large patterns. The old behaviour has been restored for atomic
1006        groups that do not contain any capturing parentheses.
1008    19. If the PCRE_NO_START_OPTIMIZE option was set for pcre_compile(), it did not
1009        suppress the check for a minimum subject length at run time. (If it was
1010        given to pcre_exec() or pcre_dfa_exec() it did work.)
1012    20. Fixed an ASCII-dependent infelicity in pcretest that would have made it
1013        fail to work when decoding hex characters in data strings in EBCDIC
1014        environments.
1016    21. It appears that in at least one Mac OS environment, the isxdigit() function
1017        is implemented as a macro that evaluates to its argument more than once,
1018        contravening the C 90 Standard (I haven't checked a later standard). There
1019        was an instance in pcretest which caused it to go wrong when processing
1020        \x{...} escapes in subject strings. The has been rewritten to avoid using
1021        things like p++ in the argument of isxdigit().
1024    Version 8.13 16-Aug-2011
1025    ------------------------
1027    1.  The Unicode data tables have been updated to Unicode 6.0.0.
1029    2.  Two minor typos in pcre_internal.h have been fixed.
1031    3.  Added #include <string.h> to pcre_scanner_unittest.cc, pcrecpp.cc, and
1032        pcrecpp_unittest.cc. They are needed for strcmp(), memset(), and strchr()
1033        in some environments (e.g. Solaris 10/SPARC using Sun Studio 12U2).
1035    4.  There were a number of related bugs in the code for matching backrefences
1036        caselessly in UTF-8 mode when codes for the characters concerned were
1037        different numbers of bytes. For example, U+023A and U+2C65 are an upper
1038        and lower case pair, using 2 and 3 bytes, respectively. The main bugs were:
1039        (a) A reference to 3 copies of a 2-byte code matched only 2 of a 3-byte
1040        code. (b) A reference to 2 copies of a 3-byte code would not match 2 of a
1041        2-byte code at the end of the subject (it thought there wasn't enough data
1042        left).
1044    5.  Comprehensive information about what went wrong is now returned by
1045        pcre_exec() and pcre_dfa_exec() when the UTF-8 string check fails, as long
1046        as the output vector has at least 2 elements. The offset of the start of
1047        the failing character and a reason code are placed in the vector.
1049    6.  When the UTF-8 string check fails for pcre_compile(), the offset that is
1050        now returned is for the first byte of the failing character, instead of the
1051        last byte inspected. This is an incompatible change, but I hope it is small
1052        enough not to be a problem. It makes the returned offset consistent with
1053        pcre_exec() and pcre_dfa_exec().
1055    7.  pcretest now gives a text phrase as well as the error number when
1056        pcre_exec() or pcre_dfa_exec() fails; if the error is a UTF-8 check
1057        failure, the offset and reason code are output.
1059    8.  When \R was used with a maximizing quantifier it failed to skip backwards
1060        over a \r\n pair if the subsequent match failed. Instead, it just skipped
1061        back over a single character (\n). This seems wrong (because it treated the
1062        two characters as a single entity when going forwards), conflicts with the
1063        documentation that \R is equivalent to (?>\r\n|\n|...etc), and makes the
1064        behaviour of \R* different to (\R)*, which also seems wrong. The behaviour
1065        has been changed.
1067    9.  Some internal refactoring has changed the processing so that the handling
1068        of the PCRE_CASELESS and PCRE_MULTILINE options is done entirely at compile
1069        time (the PCRE_DOTALL option was changed this way some time ago: version
1070        7.7 change 16). This has made it possible to abolish the OP_OPT op code,
1071        which was always a bit of a fudge. It also means that there is one less
1072        argument for the match() function, which reduces its stack requirements
1073        slightly. This change also fixes an incompatibility with Perl: the pattern
1074        (?i:([^b]))(?1) should not match "ab", but previously PCRE gave a match.
1076    10. More internal refactoring has drastically reduced the number of recursive
1077        calls to match() for possessively repeated groups such as (abc)++ when
1078        using pcre_exec().
1080    11. While implementing 10, a number of bugs in the handling of groups were
1081        discovered and fixed:
1083        (?<=(a)+) was not diagnosed as invalid (non-fixed-length lookbehind).
1084        (a|)*(?1) gave a compile-time internal error.
1085        ((a|)+)+  did not notice that the outer group could match an empty string.
1086        (^a|^)+   was not marked as anchored.
1087        (.*a|.*)+ was not marked as matching at start or after a newline.
1089    12. Yet more internal refactoring has removed another argument from the match()
1090        function. Special calls to this function are now indicated by setting a
1091        value in a variable in the "match data" data block.
1093    13. Be more explicit in pcre_study() instead of relying on "default" for
1094        opcodes that mean there is no starting character; this means that when new
1095        ones are added and accidentally left out of pcre_study(), testing should
1096        pick them up.
1098    14. The -s option of pcretest has been documented for ages as being an old
1099        synonym of -m (show memory usage). I have changed it to mean "force study
1100        for every regex", that is, assume /S for every regex. This is similar to -i
1101        and -d etc. It's slightly incompatible, but I'm hoping nobody is still
1102        using it. It makes it easier to run collections of tests with and without
1103        study enabled, and thereby test pcre_study() more easily. All the standard
1104        tests are now run with and without -s (but some patterns can be marked as
1105        "never study" - see 20 below).
1107    15. When (*ACCEPT) was used in a subpattern that was called recursively, the
1108        restoration of the capturing data to the outer values was not happening
1109        correctly.
1111    16. If a recursively called subpattern ended with (*ACCEPT) and matched an
1112        empty string, and PCRE_NOTEMPTY was set, pcre_exec() thought the whole
1113        pattern had matched an empty string, and so incorrectly returned a no
1114        match.
1116    17. There was optimizing code for the last branch of non-capturing parentheses,
1117        and also for the obeyed branch of a conditional subexpression, which used
1118        tail recursion to cut down on stack usage. Unfortunately, now that there is
1119        the possibility of (*THEN) occurring in these branches, tail recursion is
1120        no longer possible because the return has to be checked for (*THEN). These
1121        two optimizations have therefore been removed. [But see 8.20/11 above.]
1123    18. If a pattern containing \R was studied, it was assumed that \R always
1124        matched two bytes, thus causing the minimum subject length to be
1125        incorrectly computed because \R can also match just one byte.
1127    19. If a pattern containing (*ACCEPT) was studied, the minimum subject length
1128        was incorrectly computed.
1130    20. If /S is present twice on a test pattern in pcretest input, it now
1131        *disables* studying, thereby overriding the use of -s on the command line
1132        (see 14 above). This is necessary for one or two tests to keep the output
1133        identical in both cases.
1135    21. When (*ACCEPT) was used in an assertion that matched an empty string and
1136        PCRE_NOTEMPTY was set, PCRE applied the non-empty test to the assertion.
1138    22. When an atomic group that contained a capturing parenthesis was
1139        successfully matched, but the branch in which it appeared failed, the
1140        capturing was not being forgotten if a higher numbered group was later
1141        captured. For example, /(?>(a))b|(a)c/ when matching "ac" set capturing
1142        group 1 to "a", when in fact it should be unset. This applied to multi-
1143        branched capturing and non-capturing groups, repeated or not, and also to
1144        positive assertions (capturing in negative assertions does not happen
1145        in PCRE) and also to nested atomic groups.
1147    23. Add the ++ qualifier feature to pcretest, to show the remainder of the
1148        subject after a captured substring, to make it easier to tell which of a
1149        number of identical substrings has been captured.
1151    24. The way atomic groups are processed by pcre_exec() has been changed so that
1152        if they are repeated, backtracking one repetition now resets captured
1153        values correctly. For example, if ((?>(a+)b)+aabab) is matched against
1154        "aaaabaaabaabab" the value of captured group 2 is now correctly recorded as
1155        "aaa". Previously, it would have been "a". As part of this code
1156        refactoring, the way recursive calls are handled has also been changed.
1158    25. If an assertion condition captured any substrings, they were not passed
1159        back unless some other capturing happened later. For example, if
1160        (?(?=(a))a) was matched against "a", no capturing was returned.
1162    26. When studying a pattern that contained subroutine calls or assertions,
1163        the code for finding the minimum length of a possible match was handling
1164        direct recursions such as (xxx(?1)|yyy) but not mutual recursions (where
1165        group 1 called group 2 while simultaneously a separate group 2 called group
1166        1). A stack overflow occurred in this case. I have fixed this by limiting
1167        the recursion depth to 10.
1169    27. Updated RunTest.bat in the distribution to the version supplied by Tom
1170        Fortmann. This supports explicit test numbers on the command line, and has
1171        argument validation and error reporting.
1173    28. An instance of \X with an unlimited repeat could fail if at any point the
1174        first character it looked at was a mark character.
1176    29. Some minor code refactoring concerning Unicode properties and scripts
1177        should reduce the stack requirement of match() slightly.
1179    30. Added the '=' option to pcretest to check the setting of unused capturing
1180        slots at the end of the pattern, which are documented as being -1, but are
1181        not included in the return count.
1183    31. If \k was not followed by a braced, angle-bracketed, or quoted name, PCRE
1184        compiled something random. Now it gives a compile-time error (as does
1185        Perl).
1187    32. A *MARK encountered during the processing of a positive assertion is now
1188        recorded and passed back (compatible with Perl).
1190    33. If --only-matching or --colour was set on a pcregrep call whose pattern
1191        had alternative anchored branches, the search for a second match in a line
1192        was done as if at the line start. Thus, for example, /^01|^02/ incorrectly
1193        matched the line "0102" twice. The same bug affected patterns that started
1194        with a backwards assertion. For example /\b01|\b02/ also matched "0102"
1195        twice.
1197    34. Previously, PCRE did not allow quantification of assertions. However, Perl
1198        does, and because of capturing effects, quantifying parenthesized
1199        assertions may at times be useful. Quantifiers are now allowed for
1200        parenthesized assertions.
1202    35. A minor code tidy in pcre_compile() when checking options for \R usage.
1204    36. \g was being checked for fancy things in a character class, when it should
1205        just be a literal "g".
1207    37. PCRE was rejecting [:a[:digit:]] whereas Perl was not. It seems that the
1208        appearance of a nested POSIX class supersedes an apparent external class.
1209        For example, [:a[:digit:]b:] matches "a", "b", ":", or a digit. Also,
1210        unescaped square brackets may also appear as part of class names. For
1211        example, [:a[:abc]b:] gives unknown class "[:abc]b:]". PCRE now behaves
1212        more like Perl. (But see 8.20/1 above.)
1214    38. PCRE was giving an error for \N with a braced quantifier such as {1,} (this
1215        was because it thought it was \N{name}, which is not supported).
1217    39. Add minix to OS list not supporting the -S option in pcretest.
1219    40. PCRE tries to detect cases of infinite recursion at compile time, but it
1220        cannot analyze patterns in sufficient detail to catch mutual recursions
1221        such as ((?1))((?2)). There is now a runtime test that gives an error if a
1222        subgroup is called recursively as a subpattern for a second time at the
1223        same position in the subject string. In previous releases this might have
1224        been caught by the recursion limit, or it might have run out of stack.
1226    41. A pattern such as /(?(R)a+|(?R)b)/ is quite safe, as the recursion can
1227        happen only once. PCRE was, however incorrectly giving a compile time error
1228        "recursive call could loop indefinitely" because it cannot analyze the
1229        pattern in sufficient detail. The compile time test no longer happens when
1230        PCRE is compiling a conditional subpattern, but actual runaway loops are
1231        now caught at runtime (see 40 above).
1233    42. It seems that Perl allows any characters other than a closing parenthesis
1234        to be part of the NAME in (*MARK:NAME) and other backtracking verbs. PCRE
1235        has been changed to be the same.
1237    43. Updated configure.ac to put in more quoting round AC_LANG_PROGRAM etc. so
1238        as not to get warnings when autogen.sh is called. Also changed
1239        AC_PROG_LIBTOOL (deprecated) to LT_INIT (the current macro).
1241    44. To help people who use pcregrep to scan files containing exceedingly long
1242        lines, the following changes have been made:
1244        (a) The default value of the buffer size parameter has been increased from
1245            8K to 20K. (The actual buffer used is three times this size.)
1247        (b) The default can be changed by ./configure --with-pcregrep-bufsize when
1248            PCRE is built.
1250        (c) A --buffer-size=n option has been added to pcregrep, to allow the size
1251            to be set at run time.
1253        (d) Numerical values in pcregrep options can be followed by K or M, for
1254            example --buffer-size=50K.
1256        (e) If a line being scanned overflows pcregrep's buffer, an error is now
1257            given and the return code is set to 2.
1259    45. Add a pointer to the latest mark to the callout data block.
1261    46. The pattern /.(*F)/, when applied to "abc" with PCRE_PARTIAL_HARD, gave a
1262        partial match of an empty string instead of no match. This was specific to
1263        the use of ".".
1265    47. The pattern /f.*/8s, when applied to "for" with PCRE_PARTIAL_HARD, gave a
1266        complete match instead of a partial match. This bug was dependent on both
1267        the PCRE_UTF8 and PCRE_DOTALL options being set.
1269    48. For a pattern such as /\babc|\bdef/ pcre_study() was failing to set up the
1270        starting byte set, because \b was not being ignored.
1273    Version 8.12 15-Jan-2011
1274    ------------------------
1276    1.  Fixed some typos in the markup of the man pages, and wrote a script that
1277        checks for such things as part of the documentation building process.
1279    2.  On a big-endian 64-bit system, pcregrep did not correctly process the
1280        --match-limit and --recursion-limit options (added for 8.11). In
1281        particular, this made one of the standard tests fail. (The integer value
1282        went into the wrong half of a long int.)
1284    3.  If the --colour option was given to pcregrep with -v (invert match), it
1285        did strange things, either producing crazy output, or crashing. It should,
1286        of course, ignore a request for colour when reporting lines that do not
1287        match.
1289    4.  Another pcregrep bug caused similar problems if --colour was specified with
1290        -M (multiline) and the pattern match finished with a line ending.
1292    5.  In pcregrep, when a pattern that ended with a literal newline sequence was
1293        matched in multiline mode, the following line was shown as part of the
1294        match. This seems wrong, so I have changed it.
1296    6.  Another pcregrep bug in multiline mode, when --colour was specified, caused
1297        the check for further matches in the same line (so they could be coloured)
1298        to overrun the end of the current line. If another match was found, it was
1299        incorrectly shown (and then shown again when found in the next line).
1301    7.  If pcregrep was compiled under Windows, there was a reference to the
1302        function pcregrep_exit() before it was defined. I am assuming this was
1303        the cause of the "error C2371: 'pcregrep_exit' : redefinition;" that was
1304        reported by a user. I've moved the definition above the reference.
1307    Version 8.11 10-Dec-2010
1308    ------------------------
1310    1.  (*THEN) was not working properly if there were untried alternatives prior
1311        to it in the current branch. For example, in ((a|b)(*THEN)(*F)|c..) it
1312        backtracked to try for "b" instead of moving to the next alternative branch
1313        at the same level (in this case, to look for "c"). The Perl documentation
1314        is clear that when (*THEN) is backtracked onto, it goes to the "next
1315        alternative in the innermost enclosing group".
1317    2.  (*COMMIT) was not overriding (*THEN), as it does in Perl. In a pattern
1318        such as   (A(*COMMIT)B(*THEN)C|D)  any failure after matching A should
1319        result in overall failure. Similarly, (*COMMIT) now overrides (*PRUNE) and
1320        (*SKIP), (*SKIP) overrides (*PRUNE) and (*THEN), and (*PRUNE) overrides
1321        (*THEN).
1323    3.  If \s appeared in a character class, it removed the VT character from
1324        the class, even if it had been included by some previous item, for example
1325        in [\x00-\xff\s]. (This was a bug related to the fact that VT is not part
1326        of \s, but is part of the POSIX "space" class.)
1328    4.  A partial match never returns an empty string (because you can always
1329        match an empty string at the end of the subject); however the checking for
1330        an empty string was starting at the "start of match" point. This has been
1331        changed to the "earliest inspected character" point, because the returned
1332        data for a partial match starts at this character. This means that, for
1333        example, /(?<=abc)def/ gives a partial match for the subject "abc"
1334        (previously it gave "no match").
1336    5.  Changes have been made to the way PCRE_PARTIAL_HARD affects the matching
1337        of $, \z, \Z, \b, and \B. If the match point is at the end of the string,
1338        previously a full match would be given. However, setting PCRE_PARTIAL_HARD
1339        has an implication that the given string is incomplete (because a partial
1340        match is preferred over a full match). For this reason, these items now
1341        give a partial match in this situation. [Aside: previously, the one case
1342        /t\b/ matched against "cat" with PCRE_PARTIAL_HARD set did return a partial
1343        match rather than a full match, which was wrong by the old rules, but is
1344        now correct.]
1346    6.  There was a bug in the handling of #-introduced comments, recognized when
1347        PCRE_EXTENDED is set, when PCRE_NEWLINE_ANY and PCRE_UTF8 were also set.
1348        If a UTF-8 multi-byte character included the byte 0x85 (e.g. +U0445, whose
1349        UTF-8 encoding is 0xd1,0x85), this was misinterpreted as a newline when
1350        scanning for the end of the comment. (*Character* 0x85 is an "any" newline,
1351        but *byte* 0x85 is not, in UTF-8 mode). This bug was present in several
1352        places in pcre_compile().
1354    7.  Related to (6) above, when pcre_compile() was skipping #-introduced
1355        comments when looking ahead for named forward references to subpatterns,
1356        the only newline sequence it recognized was NL. It now handles newlines
1357        according to the set newline convention.
1359    8.  SunOS4 doesn't have strerror() or strtoul(); pcregrep dealt with the
1360        former, but used strtoul(), whereas pcretest avoided strtoul() but did not
1361        cater for a lack of strerror(). These oversights have been fixed.
1363    9.  Added --match-limit and --recursion-limit to pcregrep.
1365    10. Added two casts needed to build with Visual Studio when NO_RECURSE is set.
1367    11. When the -o option was used, pcregrep was setting a return code of 1, even
1368        when matches were found, and --line-buffered was not being honoured.
1370    12. Added an optional parentheses number to the -o and --only-matching options
1371        of pcregrep.
1373    13. Imitating Perl's /g action for multiple matches is tricky when the pattern
1374        can match an empty string. The code to do it in pcretest and pcredemo
1375        needed fixing:
1377        (a) When the newline convention was "crlf", pcretest got it wrong, skipping
1378            only one byte after an empty string match just before CRLF (this case
1379            just got forgotten; "any" and "anycrlf" were OK).
1381        (b) The pcretest code also had a bug, causing it to loop forever in UTF-8
1382            mode when an empty string match preceded an ASCII character followed by
1383            a non-ASCII character. (The code for advancing by one character rather
1384            than one byte was nonsense.)
1386        (c) The pcredemo.c sample program did not have any code at all to handle
1387            the cases when CRLF is a valid newline sequence.
1389    14. Neither pcre_exec() nor pcre_dfa_exec() was checking that the value given
1390        as a starting offset was within the subject string. There is now a new
1391        error, PCRE_ERROR_BADOFFSET, which is returned if the starting offset is
1392        negative or greater than the length of the string. In order to test this,
1393        pcretest is extended to allow the setting of negative starting offsets.
1395    15. In both pcre_exec() and pcre_dfa_exec() the code for checking that the
1396        starting offset points to the beginning of a UTF-8 character was
1397        unnecessarily clumsy. I tidied it up.
1399    16. Added PCRE_ERROR_SHORTUTF8 to make it possible to distinguish between a
1400        bad UTF-8 sequence and one that is incomplete when using PCRE_PARTIAL_HARD.
1402    17. Nobody had reported that the --include_dir option, which was added in
1403        release 7.7 should have been called --include-dir (hyphen, not underscore)
1404        for compatibility with GNU grep. I have changed it to --include-dir, but
1405        left --include_dir as an undocumented synonym, and the same for
1406        --exclude-dir, though that is not available in GNU grep, at least as of
1407        release 2.5.4.
1409    18. At a user's suggestion, the macros GETCHAR and friends (which pick up UTF-8
1410        characters from a string of bytes) have been redefined so as not to use
1411        loops, in order to improve performance in some environments. At the same
1412        time, I abstracted some of the common code into auxiliary macros to save
1413        repetition (this should not affect the compiled code).
1415    19. If \c was followed by a multibyte UTF-8 character, bad things happened. A
1416        compile-time error is now given if \c is not followed by an ASCII
1417        character, that is, a byte less than 128. (In EBCDIC mode, the code is
1418        different, and any byte value is allowed.)
1420    20. Recognize (*NO_START_OPT) at the start of a pattern to set the PCRE_NO_
1421        START_OPTIMIZE option, which is now allowed at compile time - but just
1422        passed through to pcre_exec() or pcre_dfa_exec(). This makes it available
1423        to pcregrep and other applications that have no direct access to PCRE
1424        options. The new /Y option in pcretest sets this option when calling
1425        pcre_compile().
1427    21. Change 18 of release 8.01 broke the use of named subpatterns for recursive
1428        back references. Groups containing recursive back references were forced to
1429        be atomic by that change, but in the case of named groups, the amount of
1430        memory required was incorrectly computed, leading to "Failed: internal
1431        error: code overflow". This has been fixed.
1433    22. Some patches to pcre_stringpiece.h, pcre_stringpiece_unittest.cc, and
1434        pcretest.c, to avoid build problems in some Borland environments.
1437    Version 8.10 25-Jun-2010
1438    ------------------------
1440    1.  Added support for (*MARK:ARG) and for ARG additions to PRUNE, SKIP, and
1441        THEN.
1443    2.  (*ACCEPT) was not working when inside an atomic group.
1445    3.  Inside a character class, \B is treated as a literal by default, but
1446        faulted if PCRE_EXTRA is set. This mimics Perl's behaviour (the -w option
1447        causes the error). The code is unchanged, but I tidied the documentation.
1449    4.  Inside a character class, PCRE always treated \R and \X as literals,
1450        whereas Perl faults them if its -w option is set. I have changed PCRE so
1451        that it faults them when PCRE_EXTRA is set.
1453    5.  Added support for \N, which always matches any character other than
1454        newline. (It is the same as "." when PCRE_DOTALL is not set.)
1456    6.  When compiling pcregrep with newer versions of gcc which may have
1457        FORTIFY_SOURCE set, several warnings "ignoring return value of 'fwrite',
1458        declared with attribute warn_unused_result" were given. Just casting the
1459        result to (void) does not stop the warnings; a more elaborate fudge is
1460        needed. I've used a macro to implement this.
1462    7.  Minor change to pcretest.c to avoid a compiler warning.
1464    8.  Added four artifical Unicode properties to help with an option to make
1465        \s etc use properties (see next item). The new properties are: Xan
1466        (alphanumeric), Xsp (Perl space), Xps (POSIX space), and Xwd (word).
1468    9.  Added PCRE_UCP to make \b, \d, \s, \w, and certain POSIX character classes
1469        use Unicode properties. (*UCP) at the start of a pattern can be used to set
1470        this option. Modified pcretest to add /W to test this facility. Added
1471        REG_UCP to make it available via the POSIX interface.
1473    10. Added --line-buffered to pcregrep.
1475    11. In UTF-8 mode, if a pattern that was compiled with PCRE_CASELESS was
1476        studied, and the match started with a letter with a code point greater than
1477        127 whose first byte was different to the first byte of the other case of
1478        the letter, the other case of this starting letter was not recognized
1479        (#976).
1481    12. If a pattern that was studied started with a repeated Unicode property
1482        test, for example, \p{Nd}+, there was the theoretical possibility of
1483        setting up an incorrect bitmap of starting bytes, but fortunately it could
1484        not have actually happened in practice until change 8 above was made (it
1485        added property types that matched character-matching opcodes).
1487    13. pcre_study() now recognizes \h, \v, and \R when constructing a bit map of
1488        possible starting bytes for non-anchored patterns.
1490    14. Extended the "auto-possessify" feature of pcre_compile(). It now recognizes
1491        \R, and also a number of cases that involve Unicode properties, both
1492        explicit and implicit when PCRE_UCP is set.
1494    15. If a repeated Unicode property match (e.g. \p{Lu}*) was used with non-UTF-8
1495        input, it could crash or give wrong results if characters with values
1496        greater than 0xc0 were present in the subject string. (Detail: it assumed
1497        UTF-8 input when processing these items.)
1499    16. Added a lot of (int) casts to avoid compiler warnings in systems where
1500        size_t is 64-bit (#991).
1502    17. Added a check for running out of memory when PCRE is compiled with
1503        --disable-stack-for-recursion (#990).
1505    18. If the last data line in a file for pcretest does not have a newline on
1506        the end, a newline was missing in the output.
1508    19. The default pcre_chartables.c file recognizes only ASCII characters (values
1509        less than 128) in its various bitmaps. However, there is a facility for
1510        generating tables according to the current locale when PCRE is compiled. It
1511        turns out that in some environments, 0x85 and 0xa0, which are Unicode space
1512        characters, are recognized by isspace() and therefore were getting set in
1513        these tables, and indeed these tables seem to approximate to ISO 8859. This
1514        caused a problem in UTF-8 mode when pcre_study() was used to create a list
1515        of bytes that can start a match. For \s, it was including 0x85 and 0xa0,
1516        which of course cannot start UTF-8 characters. I have changed the code so
1517        that only real ASCII characters (less than 128) and the correct starting
1518        bytes for UTF-8 encodings are set for characters greater than 127 when in
1519        UTF-8 mode. (When PCRE_UCP is set - see 9 above - the code is different
1520        altogether.)
1522    20. Added the /T option to pcretest so as to be able to run tests with non-
1523        standard character tables, thus making it possible to include the tests
1524        used for 19 above in the standard set of tests.
1526    21. A pattern such as (?&t)(?#()(?(DEFINE)(?<t>a)) which has a forward
1527        reference to a subpattern the other side of a comment that contains an
1528        opening parenthesis caused either an internal compiling error, or a
1529        reference to the wrong subpattern.
1532    Version 8.02 19-Mar-2010
1533    ------------------------
1535    1.  The Unicode data tables have been updated to Unicode 5.2.0.
1537    2.  Added the option --libs-cpp to pcre-config, but only when C++ support is
1538        configured.
1540    3.  Updated the licensing terms in the pcregexp.pas file, as agreed with the
1541        original author of that file, following a query about its status.
1543    4.  On systems that do not have stdint.h (e.g. Solaris), check for and include
1544        inttypes.h instead. This fixes a bug that was introduced by change 8.01/8.
1546    5.  A pattern such as (?&t)*+(?(DEFINE)(?<t>.)) which has a possessive
1547        quantifier applied to a forward-referencing subroutine call, could compile
1548        incorrect code or give the error "internal error: previously-checked
1549        referenced subpattern not found".
1551    6.  Both MS Visual Studio and Symbian OS have problems with initializing
1552        variables to point to external functions. For these systems, therefore,
1553        pcre_malloc etc. are now initialized to local functions that call the
1554        relevant global functions.
1556    7.  There were two entries missing in the vectors called coptable and poptable
1557        in pcre_dfa_exec.c. This could lead to memory accesses outsize the vectors.
1558        I've fixed the data, and added a kludgy way of testing at compile time that
1559        the lengths are correct (equal to the number of opcodes).
1561    8.  Following on from 7, I added a similar kludge to check the length of the
1562        eint vector in pcreposix.c.
1564    9.  Error texts for pcre_compile() are held as one long string to avoid too
1565        much relocation at load time. To find a text, the string is searched,
1566        counting zeros. There was no check for running off the end of the string,
1567        which could happen if a new error number was added without updating the
1568        string.
1570    10. \K gave a compile-time error if it appeared in a lookbehind assersion.
1572    11. \K was not working if it appeared in an atomic group or in a group that
1573        was called as a "subroutine", or in an assertion. Perl 5.11 documents that
1574        \K is "not well defined" if used in an assertion. PCRE now accepts it if
1575        the assertion is positive, but not if it is negative.
1577    12. Change 11 fortuitously reduced the size of the stack frame used in the
1578        "match()" function of pcre_exec.c by one pointer. Forthcoming
1579        implementation of support for (*MARK) will need an extra pointer on the
1580        stack; I have reserved it now, so that the stack frame size does not
1581        decrease.
1583    13. A pattern such as (?P<L1>(?P<L2>0)|(?P>L2)(?P>L1)) in which the only other
1584        item in branch that calls a recursion is a subroutine call - as in the
1585        second branch in the above example - was incorrectly given the compile-
1586        time error "recursive call could loop indefinitely" because pcre_compile()
1587        was not correctly checking the subroutine for matching a non-empty string.
1589    14. The checks for overrunning compiling workspace could trigger after an
1590        overrun had occurred. This is a "should never occur" error, but it can be
1591        triggered by pathological patterns such as hundreds of nested parentheses.
1592        The checks now trigger 100 bytes before the end of the workspace.
1594    15. Fix typo in configure.ac: "srtoq" should be "strtoq".
1597    Version 8.01 19-Jan-2010
1598    ------------------------
1600    1.  If a pattern contained a conditional subpattern with only one branch (in
1601        particular, this includes all (*DEFINE) patterns), a call to pcre_study()
1602        computed the wrong minimum data length (which is of course zero for such
1603        subpatterns). This could cause incorrect "no match" results.
1605    2.  For patterns such as (?i)a(?-i)b|c where an option setting at the start of
1606        the pattern is reset in the first branch, pcre_compile() failed with
1607        "internal error: code overflow at offset...". This happened only when
1608        the reset was to the original external option setting. (An optimization
1609        abstracts leading options settings into an external setting, which was the
1610        cause of this.)
1612    3.  A pattern such as ^(?!a(*SKIP)b) where a negative assertion contained one
1613        of the verbs SKIP, PRUNE, or COMMIT, did not work correctly. When the
1614        assertion pattern did not match (meaning that the assertion was true), it
1615        was incorrectly treated as false if the SKIP had been reached during the
1616        matching. This also applied to assertions used as conditions.
1618    4.  If an item that is not supported by pcre_dfa_exec() was encountered in an
1619        assertion subpattern, including such a pattern used as a condition,
1620        unpredictable results occurred, instead of the error return
1623    5.  The C++ GlobalReplace function was not working like Perl for the special
1624        situation when an empty string is matched. It now does the fancy magic
1625        stuff that is necessary.
1627    6.  In pcre_internal.h, obsolete includes to setjmp.h and stdarg.h have been
1628        removed. (These were left over from very, very early versions of PCRE.)
1630    7.  Some cosmetic changes to the code to make life easier when compiling it
1631        as part of something else:
1633        (a) Change DEBUG to PCRE_DEBUG.
1635        (b) In pcre_compile(), rename the member of the "branch_chain" structure
1636            called "current" as "current_branch", to prevent a collision with the
1637            Linux macro when compiled as a kernel module.
1639        (c) In pcre_study(), rename the function set_bit() as set_table_bit(), to
1640            prevent a collision with the Linux macro when compiled as a kernel
1641            module.
1643    8.  In pcre_compile() there are some checks for integer overflows that used to
1644        cast potentially large values to (double). This has been changed to that
1645        when building, a check for int64_t is made, and if it is found, it is used
1646        instead, thus avoiding the use of floating point arithmetic. (There is no
1647        other use of FP in PCRE.) If int64_t is not found, the fallback is to
1648        double.
1650    9.  Added two casts to avoid signed/unsigned warnings from VS Studio Express
1651        2005 (difference between two addresses compared to an unsigned value).
1653    10. Change the standard AC_CHECK_LIB test for libbz2 in configure.ac to a
1654        custom one, because of the following reported problem in Windows:
1656          - libbz2 uses the Pascal calling convention (WINAPI) for the functions
1657              under Win32.
1658          - The standard autoconf AC_CHECK_LIB fails to include "bzlib.h",
1659              therefore missing the function definition.
1660          - The compiler thus generates a "C" signature for the test function.
1661          - The linker fails to find the "C" function.
1662          - PCRE fails to configure if asked to do so against libbz2.
1664    11. When running libtoolize from libtool-2.2.6b as part of autogen.sh, these
1665        messages were output:
1667          Consider adding `AC_CONFIG_MACRO_DIR([m4])' to configure.ac and
1668          rerunning libtoolize, to keep the correct libtool macros in-tree.
1669          Consider adding `-I m4' to ACLOCAL_AMFLAGS in Makefile.am.
1671        I have done both of these things.
1673    12. Although pcre_dfa_exec() does not use nearly as much stack as pcre_exec()
1674        most of the time, it *can* run out if it is given a pattern that contains a
1675        runaway infinite recursion. I updated the discussion in the pcrestack man
1676        page.
1678    13. Now that we have gone to the x.xx style of version numbers, the minor
1679        version may start with zero. Using 08 or 09 is a bad idea because users
1680        might check the value of PCRE_MINOR in their code, and 08 or 09 may be
1681        interpreted as invalid octal numbers. I've updated the previous comment in
1682        configure.ac, and also added a check that gives an error if 08 or 09 are
1683        used.
1685    14. Change 8.00/11 was not quite complete: code had been accidentally omitted,
1686        causing partial matching to fail when the end of the subject matched \W
1687        in a UTF-8 pattern where \W was quantified with a minimum of 3.
1689    15. There were some discrepancies between the declarations in pcre_internal.h
1690        of _pcre_is_newline(), _pcre_was_newline(), and _pcre_valid_utf8() and
1691        their definitions. The declarations used "const uschar *" and the
1692        definitions used USPTR. Even though USPTR is normally defined as "const
1693        unsigned char *" (and uschar is typedeffed as "unsigned char"), it was
1694        reported that: "This difference in casting confuses some C++ compilers, for
1695        example, SunCC recognizes above declarations as different functions and
1696        generates broken code for hbpcre." I have changed the declarations to use
1697        USPTR.
1699    16. GNU libtool is named differently on some systems. The autogen.sh script now
1700        tries several variants such as glibtoolize (MacOSX) and libtoolize1x
1701        (FreeBSD).
1703    17. Applied Craig's patch that fixes an HP aCC compile error in pcre 8.00
1704        (strtoXX undefined when compiling pcrecpp.cc). The patch contains this
1705        comment: "Figure out how to create a longlong from a string: strtoll and
1706        equivalent. It's not enough to call AC_CHECK_FUNCS: hpux has a strtoll, for
1707        instance, but it only takes 2 args instead of 3!"
1709    18. A subtle bug concerned with back references has been fixed by a change of
1710        specification, with a corresponding code fix. A pattern such as
1711        ^(xa|=?\1a)+$ which contains a back reference inside the group to which it
1712        refers, was giving matches when it shouldn't. For example, xa=xaaa would
1713        match that pattern. Interestingly, Perl (at least up to 5.11.3) has the
1714        same bug. Such groups have to be quantified to be useful, or contained
1715        inside another quantified group. (If there's no repetition, the reference
1716        can never match.) The problem arises because, having left the group and
1717        moved on to the rest of the pattern, a later failure that backtracks into
1718        the group uses the captured value from the final iteration of the group
1719        rather than the correct earlier one. I have fixed this in PCRE by forcing
1720        any group that contains a reference to itself to be an atomic group; that
1721        is, there cannot be any backtracking into it once it has completed. This is
1722        similar to recursive and subroutine calls.
1725    Version 8.00 19-Oct-09
1726    ----------------------
1728    1.  The table for translating pcre_compile() error codes into POSIX error codes
1729        was out-of-date, and there was no check on the pcre_compile() error code
1730        being within the table. This could lead to an OK return being given in
1731        error.
1733    2.  Changed the call to open a subject file in pcregrep from fopen(pathname,
1734        "r") to fopen(pathname, "rb"), which fixed a problem with some of the tests
1735        in a Windows environment.
1737    3.  The pcregrep --count option prints the count for each file even when it is
1738        zero, as does GNU grep. However, pcregrep was also printing all files when
1739        --files-with-matches was added. Now, when both options are given, it prints
1740        counts only for those files that have at least one match. (GNU grep just
1741        prints the file name in this circumstance, but including the count seems
1742        more useful - otherwise, why use --count?) Also ensured that the
1743        combination -clh just lists non-zero counts, with no names.
1745    4.  The long form of the pcregrep -F option was incorrectly implemented as
1746        --fixed_strings instead of --fixed-strings. This is an incompatible change,
1747        but it seems right to fix it, and I didn't think it was worth preserving
1748        the old behaviour.
1750    5.  The command line items --regex=pattern and --regexp=pattern were not
1751        recognized by pcregrep, which required --regex pattern or --regexp pattern
1752        (with a space rather than an '='). The man page documented the '=' forms,
1753        which are compatible with GNU grep; these now work.
1755    6.  No libpcreposix.pc file was created for pkg-config; there was just
1756        libpcre.pc and libpcrecpp.pc. The omission has been rectified.
1758    7.  Added #ifndef SUPPORT_UCP into the pcre_ucd.c module, to reduce its size
1759        when UCP support is not needed, by modifying the Python script that
1760        generates it from Unicode data files. This should not matter if the module
1761        is correctly used as a library, but I received one complaint about 50K of
1762        unwanted data. My guess is that the person linked everything into his
1763        program rather than using a library. Anyway, it does no harm.
1765    8.  A pattern such as /\x{123}{2,2}+/8 was incorrectly compiled; the trigger
1766        was a minimum greater than 1 for a wide character in a possessive
1767        repetition. The same bug could also affect patterns like /(\x{ff}{0,2})*/8
1768        which had an unlimited repeat of a nested, fixed maximum repeat of a wide
1769        character. Chaos in the form of incorrect output or a compiling loop could
1770        result.
1772    9.  The restrictions on what a pattern can contain when partial matching is
1773        requested for pcre_exec() have been removed. All patterns can now be
1774        partially matched by this function. In addition, if there are at least two
1775        slots in the offset vector, the offset of the earliest inspected character
1776        for the match and the offset of the end of the subject are set in them when
1777        PCRE_ERROR_PARTIAL is returned.
1779    10. Partial matching has been split into two forms: PCRE_PARTIAL_SOFT, which is
1780        synonymous with PCRE_PARTIAL, for backwards compatibility, and
1781        PCRE_PARTIAL_HARD, which causes a partial match to supersede a full match,
1782        and may be more useful for multi-segment matching.
1784    11. Partial matching with pcre_exec() is now more intuitive. A partial match
1785        used to be given if ever the end of the subject was reached; now it is
1786        given only if matching could not proceed because another character was
1787        needed. This makes a difference in some odd cases such as Z(*FAIL) with the
1788        string "Z", which now yields "no match" instead of "partial match". In the
1789        case of pcre_dfa_exec(), "no match" is given if every matching path for the
1790        final character ended with (*FAIL).
1792    12. Restarting a match using pcre_dfa_exec() after a partial match did not work
1793        if the pattern had a "must contain" character that was already found in the
1794        earlier partial match, unless partial matching was again requested. For
1795        example, with the pattern /dog.(body)?/, the "must contain" character is
1796        "g". If the first part-match was for the string "dog", restarting with
1797        "sbody" failed. This bug has been fixed.
1799    13. The string returned by pcre_dfa_exec() after a partial match has been
1800        changed so that it starts at the first inspected character rather than the
1801        first character of the match. This makes a difference only if the pattern
1802        starts with a lookbehind assertion or \b or \B (\K is not supported by
1803        pcre_dfa_exec()). It's an incompatible change, but it makes the two
1804        matching functions compatible, and I think it's the right thing to do.
1806    14. Added a pcredemo man page, created automatically from the pcredemo.c file,
1807        so that the demonstration program is easily available in environments where
1808        PCRE has not been installed from source.
1810    15. Arranged to add -DPCRE_STATIC to cflags in libpcre.pc, libpcreposix.cp,
1811        libpcrecpp.pc and pcre-config when PCRE is not compiled as a shared
1812        library.
1814    16. Added REG_UNGREEDY to the pcreposix interface, at the request of a user.
1815        It maps to PCRE_UNGREEDY. It is not, of course, POSIX-compatible, but it
1816        is not the first non-POSIX option to be added. Clearly some people find
1817        these options useful.
1819    17. If a caller to the POSIX matching function regexec() passes a non-zero
1820        value for nmatch with a NULL value for pmatch, the value of
1821        nmatch is forced to zero.
1823    18. RunGrepTest did not have a test for the availability of the -u option of
1824        the diff command, as RunTest does. It now checks in the same way as
1825        RunTest, and also checks for the -b option.
1827    19. If an odd number of negated classes containing just a single character
1828        interposed, within parentheses, between a forward reference to a named
1829        subpattern and the definition of the subpattern, compilation crashed with
1830        an internal error, complaining that it could not find the referenced
1831        subpattern. An example of a crashing pattern is /(?&A)(([^m])(?<A>))/.
1832        [The bug was that it was starting one character too far in when skipping
1833        over the character class, thus treating the ] as data rather than
1834        terminating the class. This meant it could skip too much.]
1836    20. Added PCRE_NOTEMPTY_ATSTART in order to be able to correctly implement the
1837        /g option in pcretest when the pattern contains \K, which makes it possible
1838        to have an empty string match not at the start, even when the pattern is
1839        anchored. Updated pcretest and pcredemo to use this option.
1841    21. If the maximum number of capturing subpatterns in a recursion was greater
1842        than the maximum at the outer level, the higher number was returned, but
1843        with unset values at the outer level. The correct (outer level) value is
1844        now given.
1846    22. If (*ACCEPT) appeared inside capturing parentheses, previous releases of
1847        PCRE did not set those parentheses (unlike Perl). I have now found a way to
1848        make it do so. The string so far is captured, making this feature
1849        compatible with Perl.
1851    23. The tests have been re-organized, adding tests 11 and 12, to make it
1852        possible to check the Perl 5.10 features against Perl 5.10.
1854    24. Perl 5.10 allows subroutine calls in lookbehinds, as long as the subroutine
1855        pattern matches a fixed length string. PCRE did not allow this; now it
1856        does. Neither allows recursion.
1858    25. I finally figured out how to implement a request to provide the minimum
1859        length of subject string that was needed in order to match a given pattern.
1860        (It was back references and recursion that I had previously got hung up
1861        on.) This code has now been added to pcre_study(); it finds a lower bound
1862        to the length of subject needed. It is not necessarily the greatest lower
1863        bound, but using it to avoid searching strings that are too short does give
1864        some useful speed-ups. The value is available to calling programs via
1865        pcre_fullinfo().
1867    26. While implementing 25, I discovered to my embarrassment that pcretest had
1868        not been passing the result of pcre_study() to pcre_dfa_exec(), so the
1869        study optimizations had never been tested with that matching function.
1870        Oops. What is worse, even when it was passed study data, there was a bug in
1871        pcre_dfa_exec() that meant it never actually used it. Double oops. There
1872        were also very few tests of studied patterns with pcre_dfa_exec().
1874    27. If (?| is used to create subpatterns with duplicate numbers, they are now
1875        allowed to have the same name, even if PCRE_DUPNAMES is not set. However,
1876        on the other side of the coin, they are no longer allowed to have different
1877        names, because these cannot be distinguished in PCRE, and this has caused
1878        confusion. (This is a difference from Perl.)
1880    28. When duplicate subpattern names are present (necessarily with different
1881        numbers, as required by 27 above), and a test is made by name in a
1882        conditional pattern, either for a subpattern having been matched, or for
1883        recursion in such a pattern, all the associated numbered subpatterns are
1884        tested, and the overall condition is true if the condition is true for any
1885        one of them. This is the way Perl works, and is also more like the way
1886        testing by number works.
1889    Version 7.9 11-Apr-09
1890    ---------------------
1892    1.  When building with support for bzlib/zlib (pcregrep) and/or readline
1893        (pcretest), all targets were linked against these libraries. This included
1894        libpcre, libpcreposix, and libpcrecpp, even though they do not use these
1895        libraries. This caused unwanted dependencies to be created. This problem
1896        has been fixed, and now only pcregrep is linked with bzlib/zlib and only
1897        pcretest is linked with readline.
1899    2.  The "typedef int BOOL" in pcre_internal.h that was included inside the
1900        "#ifndef FALSE" condition by an earlier change (probably 7.8/18) has been
1901        moved outside it again, because FALSE and TRUE are already defined in AIX,
1902        but BOOL is not.
1904    3.  The pcre_config() function was treating the PCRE_MATCH_LIMIT and
1905        PCRE_MATCH_LIMIT_RECURSION values as ints, when they should be long ints.
1907    4.  The pcregrep documentation said spaces were inserted as well as colons (or
1908        hyphens) following file names and line numbers when outputting matching
1909        lines. This is not true; no spaces are inserted. I have also clarified the
1910        wording for the --colour (or --color) option.
1912    5.  In pcregrep, when --colour was used with -o, the list of matching strings
1913        was not coloured; this is different to GNU grep, so I have changed it to be
1914        the same.
1916    6.  When --colo(u)r was used in pcregrep, only the first matching substring in
1917        each matching line was coloured. Now it goes on to look for further matches
1918        of any of the test patterns, which is the same behaviour as GNU grep.
1920    7.  A pattern that could match an empty string could cause pcregrep to loop; it
1921        doesn't make sense to accept an empty string match in pcregrep, so I have
1922        locked it out (using PCRE's PCRE_NOTEMPTY option). By experiment, this
1923        seems to be how GNU grep behaves. [But see later change 40 for release
1924        8.33.]
1926    8.  The pattern (?(?=.*b)b|^) was incorrectly compiled as "match must be at
1927        start or after a newline", because the conditional assertion was not being
1928        correctly handled. The rule now is that both the assertion and what follows
1929        in the first alternative must satisfy the test.
1931    9.  If auto-callout was enabled in a pattern with a conditional group whose
1932        condition was an assertion, PCRE could crash during matching, both with
1933        pcre_exec() and pcre_dfa_exec().
1935    10. The PCRE_DOLLAR_ENDONLY option was not working when pcre_dfa_exec() was
1936        used for matching.
1938    11. Unicode property support in character classes was not working for
1939        characters (bytes) greater than 127 when not in UTF-8 mode.
1941    12. Added the -M command line option to pcretest.
1943    14. Added the non-standard REG_NOTEMPTY option to the POSIX interface.
1945    15. Added the PCRE_NO_START_OPTIMIZE match-time option.
1947    16. Added comments and documentation about mis-use of no_arg in the C++
1948        wrapper.
1950    17. Implemented support for UTF-8 encoding in EBCDIC environments, a patch
1951        from Martin Jerabek that uses macro names for all relevant character and
1952        string constants.
1954    18. Added to pcre_internal.h two configuration checks: (a) If both EBCDIC and
1955        SUPPORT_UTF8 are set, give an error; (b) If SUPPORT_UCP is set without
1956        SUPPORT_UTF8, define SUPPORT_UTF8. The "configure" script handles both of
1957        these, but not everybody uses configure.
1959    19. A conditional group that had only one branch was not being correctly
1960        recognized as an item that could match an empty string. This meant that an
1961        enclosing group might also not be so recognized, causing infinite looping
1962        (and probably a segfault) for patterns such as ^"((?(?=[a])[^"])|b)*"$
1963        with the subject "ab", where knowledge that the repeated group can match
1964        nothing is needed in order to break the loop.
1966    20. If a pattern that was compiled with callouts was matched using pcre_dfa_
1967        exec(), but without supplying a callout function, matching went wrong.
1969    21. If PCRE_ERROR_MATCHLIMIT occurred during a recursion, there was a memory
1970        leak if the size of the offset vector was greater than 30. When the vector
1971        is smaller, the saved offsets during recursion go onto a local stack
1972        vector, but for larger vectors malloc() is used. It was failing to free
1973        when the recursion yielded PCRE_ERROR_MATCH_LIMIT (or any other "abnormal"
1974        error, in fact).
1976    22. There was a missing #ifdef SUPPORT_UTF8 round one of the variables in the
1977        heapframe that is used only when UTF-8 support is enabled. This caused no
1978        problem, but was untidy.
1980    23. Steven Van Ingelgem's patch to CMakeLists.txt to change the name
1981        CMAKE_BINARY_DIR to PROJECT_BINARY_DIR so that it works when PCRE is
1982        included within another project.
1984    24. Steven Van Ingelgem's patches to add more options to the CMake support,
1985        slightly modified by me:
1987          (a) PCRE_BUILD_TESTS can be set OFF not to build the tests, including
1988              not building pcregrep.
1990          (b) PCRE_BUILD_PCREGREP can be see OFF not to build pcregrep, but only
1991              if PCRE_BUILD_TESTS is also set OFF, because the tests use pcregrep.
1993    25. Forward references, both numeric and by name, in patterns that made use of
1994        duplicate group numbers, could behave incorrectly or give incorrect errors,
1995        because when scanning forward to find the reference group, PCRE was not
1996        taking into account the duplicate group numbers. A pattern such as
1997        ^X(?3)(a)(?|(b)|(q))(Y) is an example.
1999    26. Changed a few more instances of "const unsigned char *" to USPTR, making
2000        the feature of a custom pointer more persuasive (as requested by a user).
2002    27. Wrapped the definitions of fileno and isatty for Windows, which appear in
2003        pcretest.c, inside #ifndefs, because it seems they are sometimes already
2004        pre-defined.
2006    28. Added support for (*UTF8) at the start of a pattern.
2008    29. Arrange for flags added by the "release type" setting in CMake to be shown
2009        in the configuration summary.
2012    Version 7.8 05-Sep-08
2013    ---------------------
2015    1.  Replaced UCP searching code with optimized version as implemented for Ad
2016        Muncher (http://www.admuncher.com/) by Peter Kankowski. This uses a two-
2017        stage table and inline lookup instead of a function, giving speed ups of 2
2018        to 5 times on some simple patterns that I tested. Permission was given to
2019        distribute the MultiStage2.py script that generates the tables (it's not in
2020        the tarball, but is in the Subversion repository).
2022    2.  Updated the Unicode datatables to Unicode 5.1.0. This adds yet more
2023        scripts.
2025    3.  Change 12 for 7.7 introduced a bug in pcre_study() when a pattern contained
2026        a group with a zero qualifier. The result of the study could be incorrect,
2027        or the function might crash, depending on the pattern.
2029    4.  Caseless matching was not working for non-ASCII characters in back
2030        references. For example, /(\x{de})\1/8i was not matching \x{de}\x{fe}.
2031        It now works when Unicode Property Support is available.
2033    5.  In pcretest, an escape such as \x{de} in the data was always generating
2034        a UTF-8 string, even in non-UTF-8 mode. Now it generates a single byte in
2035        non-UTF-8 mode. If the value is greater than 255, it gives a warning about
2036        truncation.
2038    6.  Minor bugfix in pcrecpp.cc (change "" == ... to NULL == ...).
2040    7.  Added two (int) casts to pcregrep when printing the difference of two
2041        pointers, in case they are 64-bit values.
2043    8.  Added comments about Mac OS X stack usage to the pcrestack man page and to
2044        test 2 if it fails.
2046    9.  Added PCRE_CALL_CONVENTION just before the names of all exported functions,
2047        and a #define of that name to empty if it is not externally set. This is to
2048        allow users of MSVC to set it if necessary.
2050    10. The PCRE_EXP_DEFN macro which precedes exported functions was missing from
2051        the convenience functions in the pcre_get.c source file.
2053    11. An option change at the start of a pattern that had top-level alternatives
2054        could cause overwriting and/or a crash. This command provoked a crash in
2055        some environments:
2057          printf "/(?i)[\xc3\xa9\xc3\xbd]|[\xc3\xa9\xc3\xbdA]/8\n" | pcretest
2059        This potential security problem was recorded as CVE-2008-2371.
2061    12. For a pattern where the match had to start at the beginning or immediately
2062        after a newline (e.g /.*anything/ without the DOTALL flag), pcre_exec() and
2063        pcre_dfa_exec() could read past the end of the passed subject if there was
2064        no match. To help with detecting such bugs (e.g. with valgrind), I modified
2065        pcretest so that it places the subject at the end of its malloc-ed buffer.
2067    13. The change to pcretest in 12 above threw up a couple more cases when pcre_
2068        exec() might read past the end of the data buffer in UTF-8 mode.
2070    14. A similar bug to 7.3/2 existed when the PCRE_FIRSTLINE option was set and
2071        the data contained the byte 0x85 as part of a UTF-8 character within its
2072        first line. This applied both to normal and DFA matching.
2074    15. Lazy qualifiers were not working in some cases in UTF-8 mode. For example,
2075        /^[^d]*?$/8 failed to match "abc".
2077    16. Added a missing copyright notice to pcrecpp_internal.h.
2079    17. Make it more clear in the documentation that values returned from
2080        pcre_exec() in ovector are byte offsets, not character counts.
2082    18. Tidied a few places to stop certain compilers from issuing warnings.
2084    19. Updated the Virtual Pascal + BCC files to compile the latest v7.7, as
2085        supplied by Stefan Weber. I made a further small update for 7.8 because
2086        there is a change of source arrangements: the pcre_searchfuncs.c module is
2087        replaced by pcre_ucd.c.
2090    Version 7.7 07-May-08
2091    ---------------------
2093    1.  Applied Craig's patch to sort out a long long problem: "If we can't convert
2094        a string to a long long, pretend we don't even have a long long." This is
2095        done by checking for the strtoq, strtoll, and _strtoi64 functions.
2097    2.  Applied Craig's patch to pcrecpp.cc to restore ABI compatibility with
2098        pre-7.6 versions, which defined a global no_arg variable instead of putting
2099        it in the RE class. (See also #8 below.)
2101    3.  Remove a line of dead code, identified by coverity and reported by Nuno
2102        Lopes.
2104    4.  Fixed two related pcregrep bugs involving -r with --include or --exclude:
2106        (1) The include/exclude patterns were being applied to the whole pathnames
2107            of files, instead of just to the final components.
2109        (2) If there was more than one level of directory, the subdirectories were
2110            skipped unless they satisfied the include/exclude conditions. This is
2111            inconsistent with GNU grep (and could even be seen as contrary to the
2112            pcregrep specification - which I improved to make it absolutely clear).
2113            The action now is always to scan all levels of directory, and just
2114            apply the include/exclude patterns to regular files.
2116    5.  Added the --include_dir and --exclude_dir patterns to pcregrep, and used
2117        --exclude_dir in the tests to avoid scanning .svn directories.
2119    6.  Applied Craig's patch to the QuoteMeta function so that it escapes the
2120        NUL character as backslash + 0 rather than backslash + NUL, because PCRE
2121        doesn't support NULs in patterns.
2123    7.  Added some missing "const"s to declarations of static tables in
2124        pcre_compile.c and pcre_dfa_exec.c.
2126    8.  Applied Craig's patch to pcrecpp.cc to fix a problem in OS X that was
2127        caused by fix #2  above. (Subsequently also a second patch to fix the
2128        first patch. And a third patch - this was a messy problem.)
2130    9.  Applied Craig's patch to remove the use of push_back().
2132    10. Applied Alan Lehotsky's patch to add REG_STARTEND support to the POSIX
2133        matching function regexec().
2135    11. Added support for the Oniguruma syntax \g<name>, \g<n>, \g'name', \g'n',
2136        which, however, unlike Perl's \g{...}, are subroutine calls, not back
2137        references. PCRE supports relative numbers with this syntax (I don't think
2138        Oniguruma does).
2140    12. Previously, a group with a zero repeat such as (...){0} was completely
2141        omitted from the compiled regex. However, this means that if the group
2142        was called as a subroutine from elsewhere in the pattern, things went wrong
2143        (an internal error was given). Such groups are now left in the compiled
2144        pattern, with a new opcode that causes them to be skipped at execution
2145        time.
2147    13. Added the PCRE_JAVASCRIPT_COMPAT option. This makes the following changes
2148        to the way PCRE behaves:
2150        (a) A lone ] character is dis-allowed (Perl treats it as data).
2152        (b) A back reference to an unmatched subpattern matches an empty string
2153            (Perl fails the current match path).
2155        (c) A data ] in a character class must be notated as \] because if the
2156            first data character in a class is ], it defines an empty class. (In
2157            Perl it is not possible to have an empty class.) The empty class []
2158            never matches; it forces failure and is equivalent to (*FAIL) or (?!).
2159            The negative empty class [^] matches any one character, independently
2160            of the DOTALL setting.
2162    14. A pattern such as /(?2)[]a()b](abc)/ which had a forward reference to a
2163        non-existent subpattern following a character class starting with ']' and
2164        containing () gave an internal compiling error instead of "reference to
2165        non-existent subpattern". Fortunately, when the pattern did exist, the
2166        compiled code was correct. (When scanning forwards to check for the
2167        existence of the subpattern, it was treating the data ']' as terminating
2168        the class, so got the count wrong. When actually compiling, the reference
2169        was subsequently set up correctly.)
2171    15. The "always fail" assertion (?!) is optimzed to (*FAIL) by pcre_compile;
2172        it was being rejected as not supported by pcre_dfa_exec(), even though
2173        other assertions are supported. I have made pcre_dfa_exec() support
2174        (*FAIL).
2176    16. The implementation of 13c above involved the invention of a new opcode,
2177        OP_ALLANY, which is like OP_ANY but doesn't check the /s flag. Since /s
2178        cannot be changed at match time, I realized I could make a small
2179        improvement to matching performance by compiling OP_ALLANY instead of
2180        OP_ANY for "." when DOTALL was set, and then removing the runtime tests
2181        on the OP_ANY path.
2183    17. Compiling pcretest on Windows with readline support failed without the
2184        following two fixes: (1) Make the unistd.h include conditional on
2185        HAVE_UNISTD_H; (2) #define isatty and fileno as _isatty and _fileno.
2187    18. Changed CMakeLists.txt and cmake/FindReadline.cmake to arrange for the
2188        ncurses library to be included for pcretest when ReadLine support is
2189        requested, but also to allow for it to be overridden. This patch came from
2190        Daniel Bergström.
2192    19. There was a typo in the file ucpinternal.h where f0_rangeflag was defined
2193        as 0x00f00000 instead of 0x00800000. Luckily, this would not have caused
2194        any errors with the current Unicode tables. Thanks to Peter Kankowski for
2195        spotting this.
2198    Version 7.6 28-Jan-08
2199    ---------------------
2201    1.  A character class containing a very large number of characters with
2202        codepoints greater than 255 (in UTF-8 mode, of course) caused a buffer
2203        overflow.
2205    2.  Patch to cut out the "long long" test in pcrecpp_unittest when
2206        HAVE_LONG_LONG is not defined.
2208    3.  Applied Christian Ehrlicher's patch to update the CMake build files to
2209        bring them up to date and include new features. This patch includes:
2211        - Fixed PH's badly added libz and libbz2 support.
2212        - Fixed a problem with static linking.
2213        - Added pcredemo. [But later removed - see 7 below.]
2214        - Fixed dftables problem and added an option.
2215        - Added a number of HAVE_XXX tests, including HAVE_WINDOWS_H and
2216            HAVE_LONG_LONG.
2217        - Added readline support for pcretest.
2218        - Added an listing of the option settings after cmake has run.
2220    4.  A user submitted a patch to Makefile that makes it easy to create
2221        "pcre.dll" under mingw when using Configure/Make. I added stuff to
2222        Makefile.am that cause it to include this special target, without
2223        affecting anything else. Note that the same mingw target plus all
2224        the other distribution libraries and programs are now supported
2225        when configuring with CMake (see 6 below) instead of with
2226        Configure/Make.
2228    5.  Applied Craig's patch that moves no_arg into the RE class in the C++ code.
2229        This is an attempt to solve the reported problem "pcrecpp::no_arg is not
2230        exported in the Windows port". It has not yet been confirmed that the patch
2231        solves the problem, but it does no harm.
2233    6.  Applied Sheri's patch to CMakeLists.txt to add NON_STANDARD_LIB_PREFIX and
2234        NON_STANDARD_LIB_SUFFIX for dll names built with mingw when configured
2235        with CMake, and also correct the comment about stack recursion.
2237    7.  Remove the automatic building of pcredemo from the ./configure system and
2238        from CMakeLists.txt. The whole idea of pcredemo.c is that it is an example
2239        of a program that users should build themselves after PCRE is installed, so
2240        building it automatically is not really right. What is more, it gave
2241        trouble in some build environments.
2243    8.  Further tidies to CMakeLists.txt from Sheri and Christian.
2246    Version 7.5 10-Jan-08
2247    ---------------------
2249    1.  Applied a patch from Craig: "This patch makes it possible to 'ignore'
2250        values in parens when parsing an RE using the C++ wrapper."
2252    2.  Negative specials like \S did not work in character classes in UTF-8 mode.
2253        Characters greater than 255 were excluded from the class instead of being
2254        included.
2256    3.  The same bug as (2) above applied to negated POSIX classes such as
2257        [:^space:].
2259    4.  PCRECPP_STATIC was referenced in pcrecpp_internal.h, but nowhere was it
2260        defined or documented. It seems to have been a typo for PCRE_STATIC, so
2261        I have changed it.
2263    5.  The construct (?&) was not diagnosed as a syntax error (it referenced the
2264        first named subpattern) and a construct such as (?&a) would reference the
2265        first named subpattern whose name started with "a" (in other words, the
2266        length check was missing). Both these problems are fixed. "Subpattern name
2267        expected" is now given for (?&) (a zero-length name), and this patch also
2268        makes it give the same error for \k'' (previously it complained that that
2269        was a reference to a non-existent subpattern).
2271    6.  The erroneous patterns (?+-a) and (?-+a) give different error messages;
2272        this is right because (?- can be followed by option settings as well as by
2273        digits. I have, however, made the messages clearer.
2275    7.  Patterns such as (?(1)a|b) (a pattern that contains fewer subpatterns
2276        than the number used in the conditional) now cause a compile-time error.
2277        This is actually not compatible with Perl, which accepts such patterns, but
2278        treats the conditional as always being FALSE (as PCRE used to), but it
2279        seems to me that giving a diagnostic is better.
2281    8.  Change "alphameric" to the more common word "alphanumeric" in comments
2282        and messages.
2284    9.  Fix two occurrences of "backslash" in comments that should have been
2285        "backspace".
2287    10. Remove two redundant lines of code that can never be obeyed (their function
2288        was moved elsewhere).
2290    11. The program that makes PCRE's Unicode character property table had a bug
2291        which caused it to generate incorrect table entries for sequences of
2292        characters that have the same character type, but are in different scripts.
2293        It amalgamated them into a single range, with the script of the first of
2294        them. In other words, some characters were in the wrong script. There were
2295        thirteen such cases, affecting characters in the following ranges:
2297          U+002b0 - U+002c1
2298          U+0060c - U+0060d
2299          U+0061e - U+00612
2300          U+0064b - U+0065e
2301          U+0074d - U+0076d
2302          U+01800 - U+01805
2303          U+01d00 - U+01d77
2304          U+01d9b - U+01dbf
2305          U+0200b - U+0200f
2306          U+030fc - U+030fe
2307          U+03260 - U+0327f
2308          U+0fb46 - U+0fbb1
2309          U+10450 - U+1049d
2311    12. The -o option (show only the matching part of a line) for pcregrep was not
2312        compatible with GNU grep in that, if there was more than one match in a
2313        line, it showed only the first of them. It now behaves in the same way as
2314        GNU grep.
2316    13. If the -o and -v options were combined for pcregrep, it printed a blank
2317        line for every non-matching line. GNU grep prints nothing, and pcregrep now
2318        does the same. The return code can be used to tell if there were any
2319        non-matching lines.
2321    14. Added --file-offsets and --line-offsets to pcregrep.
2323    15. The pattern (?=something)(?R) was not being diagnosed as a potentially
2324        infinitely looping recursion. The bug was that positive lookaheads were not
2325        being skipped when checking for a possible empty match (negative lookaheads
2326        and both kinds of lookbehind were skipped).
2328    16. Fixed two typos in the Windows-only code in pcregrep.c, and moved the
2329        inclusion of <windows.h> to before rather than after the definition of
2330        INVALID_FILE_ATTRIBUTES (patch from David Byron).
2332    17. Specifying a possessive quantifier with a specific limit for a Unicode
2333        character property caused pcre_compile() to compile bad code, which led at
2334        runtime to PCRE_ERROR_INTERNAL (-14). Examples of patterns that caused this
2335        are: /\p{Zl}{2,3}+/8 and /\p{Cc}{2}+/8. It was the possessive "+" that
2336        caused the error; without that there was no problem.
2338    18. Added --enable-pcregrep-libz and --enable-pcregrep-libbz2.
2340    19. Added --enable-pcretest-libreadline.
2342    20. In pcrecpp.cc, the variable 'count' was incremented twice in
2343        RE::GlobalReplace(). As a result, the number of replacements returned was
2344        double what it should be. I removed one of the increments, but Craig sent a
2345        later patch that removed the other one (the right fix) and added unit tests
2346        that check the return values (which was not done before).
2348    21. Several CMake things:
2350        (1) Arranged that, when cmake is used on Unix, the libraries end up with
2351            the names libpcre and libpcreposix, not just pcre and pcreposix.
2353        (2) The above change means that pcretest and pcregrep are now correctly
2354            linked with the newly-built libraries, not previously installed ones.
2358    22. In UTF-8 mode, with newline set to "any", a pattern such as .*a.*=.b.*
2359        crashed when matching a string such as a\x{2029}b (note that \x{2029} is a
2360        UTF-8 newline character). The key issue is that the pattern starts .*;
2361        this means that the match must be either at the beginning, or after a
2362        newline. The bug was in the code for advancing after a failed match and
2363        checking that the new position followed a newline. It was not taking
2364        account of UTF-8 characters correctly.
2366    23. PCRE was behaving differently from Perl in the way it recognized POSIX
2367        character classes. PCRE was not treating the sequence [:...:] as a
2368        character class unless the ... were all letters. Perl, however, seems to
2369        allow any characters between [: and :], though of course it rejects as
2370        unknown any "names" that contain non-letters, because all the known class
2371        names consist only of letters. Thus, Perl gives an error for [[:1234:]],
2372        for example, whereas PCRE did not - it did not recognize a POSIX character
2373        class. This seemed a bit dangerous, so the code has been changed to be
2374        closer to Perl. The behaviour is not identical to Perl, because PCRE will
2375        diagnose an unknown class for, for example, [[:l\ower:]] where Perl will
2376        treat it as [[:lower:]]. However, PCRE does now give "unknown" errors where
2377        Perl does, and where it didn't before.
2379    24. Rewrite so as to remove the single use of %n from pcregrep because in some
2380        Windows environments %n is disabled by default.
2383    Version 7.4 21-Sep-07
2384    ---------------------
2386    1.  Change 7.3/28 was implemented for classes by looking at the bitmap. This
2387        means that a class such as [\s] counted as "explicit reference to CR or
2388        LF". That isn't really right - the whole point of the change was to try to
2389        help when there was an actual mention of one of the two characters. So now
2390        the change happens only if \r or \n (or a literal CR or LF) character is
2391        encountered.
2393    2.  The 32-bit options word was also used for 6 internal flags, but the numbers
2394        of both had grown to the point where there were only 3 bits left.
2395        Fortunately, there was spare space in the data structure, and so I have
2396        moved the internal flags into a new 16-bit field to free up more option
2397        bits.
2399    3.  The appearance of (?J) at the start of a pattern set the DUPNAMES option,
2400        but did not set the internal JCHANGED flag - either of these is enough to
2401        control the way the "get" function works - but the PCRE_INFO_JCHANGED
2402        facility is supposed to tell if (?J) was ever used, so now (?J) at the
2403        start sets both bits.
2405    4.  Added options (at build time, compile time, exec time) to change \R from
2406        matching any Unicode line ending sequence to just matching CR, LF, or CRLF.
2408    5.  doc/pcresyntax.html was missing from the distribution.
2410    6.  Put back the definition of PCRE_ERROR_NULLWSLIMIT, for backward
2411        compatibility, even though it is no longer used.
2413    7.  Added macro for snprintf to pcrecpp_unittest.cc and also for strtoll and
2414        strtoull to pcrecpp.cc to select the available functions in WIN32 when the
2415        windows.h file is present (where different names are used). [This was
2416        reversed later after testing - see 16 below.]
2418    8.  Changed all #include <config.h> to #include "config.h". There were also
2419        some further <pcre.h> cases that I changed to "pcre.h".
2421    9.  When pcregrep was used with the --colour option, it missed the line ending
2422        sequence off the lines that it output.
2424    10. It was pointed out to me that arrays of string pointers cause lots of
2425        relocations when a shared library is dynamically loaded. A technique of
2426        using a single long string with a table of offsets can drastically reduce
2427        these. I have refactored PCRE in four places to do this. The result is
2428        dramatic:
2430          Originally:                          290
2431          After changing UCP table:            187
2432          After changing error message table:   43
2433          After changing table of "verbs"       36
2434          After changing table of Posix names   22
2436        Thanks to the folks working on Gregex for glib for this insight.
2438    11. --disable-stack-for-recursion caused compiling to fail unless -enable-
2439        unicode-properties was also set.
2441    12. Updated the tests so that they work when \R is defaulted to ANYCRLF.
2443    13. Added checks for ANY and ANYCRLF to pcrecpp.cc where it previously
2444        checked only for CRLF.
2446    14. Added casts to pcretest.c to avoid compiler warnings.
2448    15. Added Craig's patch to various pcrecpp modules to avoid compiler warnings.
2450    16. Added Craig's patch to remove the WINDOWS_H tests, that were not working,
2451        and instead check for _strtoi64 explicitly, and avoid the use of snprintf()
2452        entirely. This removes changes made in 7 above.
2454    17. The CMake files have been updated, and there is now more information about
2455        building with CMake in the NON-UNIX-USE document.
2458    Version 7.3 28-Aug-07
2459    ---------------------
2461     1. In the rejigging of the build system that eventually resulted in 7.1, the
2462        line "#include <pcre.h>" was included in pcre_internal.h. The use of angle
2463        brackets there is not right, since it causes compilers to look for an
2464        installed pcre.h, not the version that is in the source that is being
2465        compiled (which of course may be different). I have changed it back to:
2467          #include "pcre.h"
2469        I have a vague recollection that the change was concerned with compiling in
2470        different directories, but in the new build system, that is taken care of
2471        by the VPATH setting the Makefile.
2473     2. The pattern .*$ when run in not-DOTALL UTF-8 mode with newline=any failed
2474        when the subject happened to end in the byte 0x85 (e.g. if the last
2475        character was \x{1ec5}). *Character* 0x85 is one of the "any" newline
2476        characters but of course it shouldn't be taken as a newline when it is part
2477        of another character. The bug was that, for an unlimited repeat of . in
2478        not-DOTALL UTF-8 mode, PCRE was advancing by bytes rather than by
2479        characters when looking for a newline.
2481     3. A small performance improvement in the DOTALL UTF-8 mode .* case.
2483     4. Debugging: adjusted the names of opcodes for different kinds of parentheses
2484        in debug output.
2486     5. Arrange to use "%I64d" instead of "%lld" and "%I64u" instead of "%llu" for
2487        long printing in the pcrecpp unittest when running under MinGW.
2489     6. ESC_K was left out of the EBCDIC table.
2491     7. Change 7.0/38 introduced a new limit on the number of nested non-capturing
2492        parentheses; I made it 1000, which seemed large enough. Unfortunately, the
2493        limit also applies to "virtual nesting" when a pattern is recursive, and in
2494        this case 1000 isn't so big. I have been able to remove this limit at the
2495        expense of backing off one optimization in certain circumstances. Normally,
2496        when pcre_exec() would call its internal match() function recursively and
2497        immediately return the result unconditionally, it uses a "tail recursion"
2498        feature to save stack. However, when a subpattern that can match an empty
2499        string has an unlimited repetition quantifier, it no longer makes this
2500        optimization. That gives it a stack frame in which to save the data for
2501        checking that an empty string has been matched. Previously this was taken
2502        from the 1000-entry workspace that had been reserved. So now there is no
2503        explicit limit, but more stack is used.
2505     8. Applied Daniel's patches to solve problems with the import/export magic
2506        syntax that is required for Windows, and which was going wrong for the
2507        pcreposix and pcrecpp parts of the library. These were overlooked when this
2508        problem was solved for the main library.
2510     9. There were some crude static tests to avoid integer overflow when computing
2511        the size of patterns that contain repeated groups with explicit upper
2512        limits. As the maximum quantifier is 65535, the maximum group length was
2513        set at 30,000 so that the product of these two numbers did not overflow a
2514        32-bit integer. However, it turns out that people want to use groups that
2515        are longer than 30,000 bytes (though not repeat them that many times).
2516        Change 7.0/17 (the refactoring of the way the pattern size is computed) has
2517        made it possible to implement the integer overflow checks in a much more
2518        dynamic way, which I have now done. The artificial limitation on group
2519        length has been removed - we now have only the limit on the total length of
2520        the compiled pattern, which depends on the LINK_SIZE setting.
2522    10. Fixed a bug in the documentation for get/copy named substring when
2523        duplicate names are permitted. If none of the named substrings are set, the
2524        functions return PCRE_ERROR_NOSUBSTRING (7); the doc said they returned an
2525        empty string.
2527    11. Because Perl interprets \Q...\E at a high level, and ignores orphan \E
2528        instances, patterns such as [\Q\E] or [\E] or even [^\E] cause an error,
2529        because the ] is interpreted as the first data character and the
2530        terminating ] is not found. PCRE has been made compatible with Perl in this
2531        regard. Previously, it interpreted [\Q\E] as an empty class, and [\E] could
2532        cause memory overwriting.
2534    10. Like Perl, PCRE automatically breaks an unlimited repeat after an empty
2535        string has been matched (to stop an infinite loop). It was not recognizing
2536        a conditional subpattern that could match an empty string if that
2537        subpattern was within another subpattern. For example, it looped when
2538        trying to match  (((?(1)X|))*)  but it was OK with  ((?(1)X|)*)  where the
2539        condition was not nested. This bug has been fixed.
2541    12. A pattern like \X?\d or \P{L}?\d in non-UTF-8 mode could cause a backtrack
2542        past the start of the subject in the presence of bytes with the top bit
2543        set, for example "\x8aBCD".
2545    13. Added Perl 5.10 experimental backtracking controls (*FAIL), (*F), (*PRUNE),
2546        (*SKIP), (*THEN), (*COMMIT), and (*ACCEPT).
2548    14. Optimized (?!) to (*FAIL).
2550    15. Updated the test for a valid UTF-8 string to conform to the later RFC 3629.
2551        This restricts code points to be within the range 0 to 0x10FFFF, excluding
2552        the "low surrogate" sequence 0xD800 to 0xDFFF. Previously, PCRE allowed the
2553        full range 0 to 0x7FFFFFFF, as defined by RFC 2279. Internally, it still
2554        does: it's just the validity check that is more restrictive.
2556    16. Inserted checks for integer overflows during escape sequence (backslash)
2557        processing, and also fixed erroneous offset values for syntax errors during
2558        backslash processing.
2560    17. Fixed another case of looking too far back in non-UTF-8 mode (cf 12 above)
2561        for patterns like [\PPP\x8a]{1,}\x80 with the subject "A\x80".
2563    18. An unterminated class in a pattern like (?1)\c[ with a "forward reference"
2564        caused an overrun.
2566    19. A pattern like (?:[\PPa*]*){8,} which had an "extended class" (one with
2567        something other than just ASCII characters) inside a group that had an
2568        unlimited repeat caused a loop at compile time (while checking to see
2569        whether the group could match an empty string).
2571    20. Debugging a pattern containing \p or \P could cause a crash. For example,
2572        [\P{Any}] did so. (Error in the code for printing property names.)
2574    21. An orphan \E inside a character class could cause a crash.
2576    22. A repeated capturing bracket such as (A)? could cause a wild memory
2577        reference during compilation.
2579    23. There are several functions in pcre_compile() that scan along a compiled
2580        expression for various reasons (e.g. to see if it's fixed length for look
2581        behind). There were bugs in these functions when a repeated \p or \P was
2582        present in the pattern. These operators have additional parameters compared
2583        with \d, etc, and these were not being taken into account when moving along
2584        the compiled data. Specifically:
2586        (a) A item such as \p{Yi}{3} in a lookbehind was not treated as fixed
2587            length.
2589        (b) An item such as \pL+ within a repeated group could cause crashes or
2590            loops.
2592        (c) A pattern such as \p{Yi}+(\P{Yi}+)(?1) could give an incorrect
2593            "reference to non-existent subpattern" error.
2595        (d) A pattern like (\P{Yi}{2}\277)? could loop at compile time.
2597    24. A repeated \S or \W in UTF-8 mode could give wrong answers when multibyte
2598        characters were involved (for example /\S{2}/8g with "A\x{a3}BC").
2600    25. Using pcregrep in multiline, inverted mode (-Mv) caused it to loop.
2602    26. Patterns such as [\P{Yi}A] which include \p or \P and just one other
2603        character were causing crashes (broken optimization).
2605    27. Patterns such as (\P{Yi}*\277)* (group with possible zero repeat containing
2606        \p or \P) caused a compile-time loop.
2608    28. More problems have arisen in unanchored patterns when CRLF is a valid line
2609        break. For example, the unstudied pattern [\r\n]A does not match the string
2610        "\r\nA" because change 7.0/46 below moves the current point on by two
2611        characters after failing to match at the start. However, the pattern \nA
2612        *does* match, because it doesn't start till \n, and if [\r\n]A is studied,
2613        the same is true. There doesn't seem any very clean way out of this, but
2614        what I have chosen to do makes the common cases work: PCRE now takes note
2615        of whether there can be an explicit match for \r or \n anywhere in the
2616        pattern, and if so, 7.0/46 no longer applies. As part of this change,
2617        there's a new PCRE_INFO_HASCRORLF option for finding out whether a compiled
2618        pattern has explicit CR or LF references.
2620    29. Added (*CR) etc for changing newline setting at start of pattern.
2623    Version 7.2 19-Jun-07
2624    ---------------------
2626     1. If the fr_FR locale cannot be found for test 3, try the "french" locale,
2627        which is apparently normally available under Windows.
2629     2. Re-jig the pcregrep tests with different newline settings in an attempt
2630        to make them independent of the local environment's newline setting.
2632     3. Add code to configure.ac to remove -g from the CFLAGS default settings.
2634     4. Some of the "internals" tests were previously cut out when the link size
2635        was not 2, because the output contained actual offsets. The recent new
2636        "Z" feature of pcretest means that these can be cut out, making the tests
2637        usable with all link sizes.
2639     5. Implemented Stan Switzer's goto replacement for longjmp() when not using
2640        stack recursion. This gives a massive performance boost under BSD, but just
2641        a small improvement under Linux. However, it saves one field in the frame
2642        in all cases.
2644     6. Added more features from the forthcoming Perl 5.10:
2646        (a) (?-n) (where n is a string of digits) is a relative subroutine or
2647            recursion call. It refers to the nth most recently opened parentheses.
2649        (b) (?+n) is also a relative subroutine call; it refers to the nth next
2650            to be opened parentheses.
2652        (c) Conditions that refer to capturing parentheses can be specified
2653            relatively, for example, (?(-2)... or (?(+3)...
2655        (d) \K resets the start of the current match so that everything before
2656            is not part of it.
2658        (e) \k{name} is synonymous with \k<name> and \k'name' (.NET compatible).
2660        (f) \g{name} is another synonym - part of Perl 5.10's unification of
2661            reference syntax.
2663        (g) (?| introduces a group in which the numbering of parentheses in each
2664            alternative starts with the same number.
2666        (h) \h, \H, \v, and \V match horizontal and vertical whitespace.
2668     7. Added two new calls to pcre_fullinfo(): PCRE_INFO_OKPARTIAL and
2671     8. A pattern such as  (.*(.)?)*  caused pcre_exec() to fail by either not
2672        terminating or by crashing. Diagnosed by Viktor Griph; it was in the code
2673        for detecting groups that can match an empty string.
2675     9. A pattern with a very large number of alternatives (more than several
2676        hundred) was running out of internal workspace during the pre-compile
2677        phase, where pcre_compile() figures out how much memory will be needed. A
2678        bit of new cunning has reduced the workspace needed for groups with
2679        alternatives. The 1000-alternative test pattern now uses 12 bytes of
2680        workspace instead of running out of the 4096 that are available.
2682    10. Inserted some missing (unsigned int) casts to get rid of compiler warnings.
2684    11. Applied patch from Google to remove an optimization that didn't quite work.
2685        The report of the bug said:
2687          pcrecpp::RE("a*").FullMatch("aaa") matches, while
2688          pcrecpp::RE("a*?").FullMatch("aaa") does not, and
2689          pcrecpp::RE("a*?\\z").FullMatch("aaa") does again.
2691    12. If \p or \P was used in non-UTF-8 mode on a character greater than 127
2692        it matched the wrong number of bytes.
2695    Version 7.1 24-Apr-07
2696    ---------------------
2698     1. Applied Bob Rossi and Daniel G's patches to convert the build system to one
2699        that is more "standard", making use of automake and other Autotools. There
2700        is some re-arrangement of the files and adjustment of comments consequent
2701        on this.
2703     2. Part of the patch fixed a problem with the pcregrep tests. The test of -r
2704        for recursive directory scanning broke on some systems because the files
2705        are not scanned in any specific order and on different systems the order
2706        was different. A call to "sort" has been inserted into RunGrepTest for the
2707        approprate test as a short-term fix. In the longer term there may be an
2708        alternative.
2710     3. I had an email from Eric Raymond about problems translating some of PCRE's
2711        man pages to HTML (despite the fact that I distribute HTML pages, some
2712        people do their own conversions for various reasons). The problems
2713        concerned the use of low-level troff macros .br and .in. I have therefore
2714        removed all such uses from the man pages (some were redundant, some could
2715        be replaced by .nf/.fi pairs). The 132html script that I use to generate
2716        HTML has been updated to handle .nf/.fi and to complain if it encounters
2717        .br or .in.
2719     4. Updated comments in configure.ac that get placed in config.h.in and also
2720        arranged for config.h to be included in the distribution, with the name
2721        config.h.generic, for the benefit of those who have to compile without
2722        Autotools (compare pcre.h, which is now distributed as pcre.h.generic).
2724     5. Updated the support (such as it is) for Virtual Pascal, thanks to Stefan
2725        Weber: (1) pcre_internal.h was missing some function renames; (2) updated
2726        makevp.bat for the current PCRE, using the additional files
2727        makevp_c.txt, makevp_l.txt, and pcregexp.pas.
2729     6. A Windows user reported a minor discrepancy with test 2, which turned out
2730        to be caused by a trailing space on an input line that had got lost in his
2731        copy. The trailing space was an accident, so I've just removed it.
2733     7. Add -Wl,-R... flags in pcre-config.in for *BSD* systems, as I'm told
2734        that is needed.
2736     8. Mark ucp_table (in ucptable.h) and ucp_gentype (in pcre_ucp_searchfuncs.c)
2737        as "const" (a) because they are and (b) because it helps the PHP
2738        maintainers who have recently made a script to detect big data structures
2739        in the php code that should be moved to the .rodata section. I remembered
2740        to update Builducptable as well, so it won't revert if ucptable.h is ever
2741        re-created.
2743     9. Added some extra #ifdef SUPPORT_UTF8 conditionals into pcretest.c,
2744        pcre_printint.src, pcre_compile.c, pcre_study.c, and pcre_tables.c, in
2745        order to be able to cut out the UTF-8 tables in the latter when UTF-8
2746        support is not required. This saves 1.5-2K of code, which is important in
2747        some applications.
2749        Later: more #ifdefs are needed in pcre_ord2utf8.c and pcre_valid_utf8.c
2750        so as not to refer to the tables, even though these functions will never be
2751        called when UTF-8 support is disabled. Otherwise there are problems with a
2752        shared library.
2754    10. Fixed two bugs in the emulated memmove() function in pcre_internal.h:
2756        (a) It was defining its arguments as char * instead of void *.
2758        (b) It was assuming that all moves were upwards in memory; this was true
2759            a long time ago when I wrote it, but is no longer the case.
2761        The emulated memove() is provided for those environments that have neither
2762        memmove() nor bcopy(). I didn't think anyone used it these days, but that
2763        is clearly not the case, as these two bugs were recently reported.
2765    11. The script PrepareRelease is now distributed: it calls 132html, CleanTxt,
2766        and Detrail to create the HTML documentation, the .txt form of the man
2767        pages, and it removes trailing spaces from listed files. It also creates
2768        pcre.h.generic and config.h.generic from pcre.h and config.h. In the latter
2769        case, it wraps all the #defines with #ifndefs. This script should be run
2770        before "make dist".
2772    12. Fixed two fairly obscure bugs concerned with quantified caseless matching
2773        with Unicode property support.
2775        (a) For a maximizing quantifier, if the two different cases of the
2776            character were of different lengths in their UTF-8 codings (there are
2777            some cases like this - I found 11), and the matching function had to
2778            back up over a mixture of the two cases, it incorrectly assumed they
2779            were both the same length.
2781        (b) When PCRE was configured to use the heap rather than the stack for
2782            recursion during matching, it was not correctly preserving the data for
2783            the other case of a UTF-8 character when checking ahead for a match
2784            while processing a minimizing repeat. If the check also involved
2785            matching a wide character, but failed, corruption could cause an
2786            erroneous result when trying to check for a repeat of the original
2787            character.
2789    13. Some tidying changes to the testing mechanism:
2791        (a) The RunTest script now detects the internal link size and whether there
2792            is UTF-8 and UCP support by running ./pcretest -C instead of relying on
2793            values substituted by "configure". (The RunGrepTest script already did
2794            this for UTF-8.) The configure.ac script no longer substitutes the
2795            relevant variables.
2797        (b) The debugging options /B and /D in pcretest show the compiled bytecode
2798            with length and offset values. This means that the output is different
2799            for different internal link sizes. Test 2 is skipped for link sizes
2800            other than 2 because of this, bypassing the problem. Unfortunately,
2801            there was also a test in test 3 (the locale tests) that used /B and
2802            failed for link sizes other than 2. Rather than cut the whole test out,
2803            I have added a new /Z option to pcretest that replaces the length and
2804            offset values with spaces. This is now used to make test 3 independent
2805            of link size. (Test 2 will be tidied up later.)
2807    14. If erroroffset was passed as NULL to pcre_compile, it provoked a
2808        segmentation fault instead of returning the appropriate error message.
2810    15. In multiline mode when the newline sequence was set to "any", the pattern
2811        ^$ would give a match between the \r and \n of a subject such as "A\r\nB".
2812        This doesn't seem right; it now treats the CRLF combination as the line
2813        ending, and so does not match in that case. It's only a pattern such as ^$
2814        that would hit this one: something like ^ABC$ would have failed after \r
2815        and then tried again after \r\n.
2817    16. Changed the comparison command for RunGrepTest from "diff -u" to "diff -ub"
2818        in an attempt to make files that differ only in their line terminators
2819        compare equal. This works on Linux.
2821    17. Under certain error circumstances pcregrep might try to free random memory
2822        as it exited. This is now fixed, thanks to valgrind.
2824    19. In pcretest, if the pattern /(?m)^$/g<any> was matched against the string
2825        "abc\r\n\r\n", it found an unwanted second match after the second \r. This
2826        was because its rules for how to advance for /g after matching an empty
2827        string at the end of a line did not allow for this case. They now check for
2828        it specially.
2830    20. pcretest is supposed to handle patterns and data of any length, by
2831        extending its buffers when necessary. It was getting this wrong when the
2832        buffer for a data line had to be extended.
2834    21. Added PCRE_NEWLINE_ANYCRLF which is like ANY, but matches only CR, LF, or
2835        CRLF as a newline sequence.
2837    22. Code for handling Unicode properties in pcre_dfa_exec() wasn't being cut
2838        out by #ifdef SUPPORT_UCP. This did no harm, as it could never be used, but
2839        I have nevertheless tidied it up.
2841    23. Added some casts to kill warnings from HP-UX ia64 compiler.
2843    24. Added a man page for pcre-config.
2846    Version 7.0 19-Dec-06
2847    ---------------------
2849     1. Fixed a signed/unsigned compiler warning in pcre_compile.c, shown up by
2850        moving to gcc 4.1.1.
2852     2. The -S option for pcretest uses setrlimit(); I had omitted to #include
2853        sys/time.h, which is documented as needed for this function. It doesn't
2854        seem to matter on Linux, but it showed up on some releases of OS X.
2856     3. It seems that there are systems where bytes whose values are greater than
2857        127 match isprint() in the "C" locale. The "C" locale should be the
2858        default when a C program starts up. In most systems, only ASCII printing
2859        characters match isprint(). This difference caused the output from pcretest
2860        to vary, making some of the tests fail. I have changed pcretest so that:
2862        (a) When it is outputting text in the compiled version of a pattern, bytes
2863            other than 32-126 are always shown as hex escapes.
2865        (b) When it is outputting text that is a matched part of a subject string,
2866            it does the same, unless a different locale has been set for the match
2867            (using the /L modifier). In this case, it uses isprint() to decide.
2869     4. Fixed a major bug that caused incorrect computation of the amount of memory
2870        required for a compiled pattern when options that changed within the
2871        pattern affected the logic of the preliminary scan that determines the
2872        length. The relevant options are -x, and -i in UTF-8 mode. The result was
2873        that the computed length was too small. The symptoms of this bug were
2874        either the PCRE error "internal error: code overflow" from pcre_compile(),
2875        or a glibc crash with a message such as "pcretest: free(): invalid next
2876        size (fast)". Examples of patterns that provoked this bug (shown in
2877        pcretest format) are:
2879          /(?-x: )/x
2880          /(?x)(?-x: \s*#\s*)/
2881          /((?i)[\x{c0}])/8
2882          /(?i:[\x{c0}])/8
2884        HOWEVER: Change 17 below makes this fix obsolete as the memory computation
2885        is now done differently.
2887     5. Applied patches from Google to: (a) add a QuoteMeta function to the C++
2888        wrapper classes; (b) implement a new function in the C++ scanner that is
2889        more efficient than the old way of doing things because it avoids levels of
2890        recursion in the regex matching; (c) add a paragraph to the documentation
2891        for the FullMatch() function.
2893     6. The escape sequence \n was being treated as whatever was defined as
2894        "newline". Not only was this contrary to the documentation, which states
2895        that \n is character 10 (hex 0A), but it also went horribly wrong when
2896        "newline" was defined as CRLF. This has been fixed.
2898     7. In pcre_dfa_exec.c the value of an unsigned integer (the variable called c)
2899        was being set to -1 for the "end of line" case (supposedly a value that no
2900        character can have). Though this value is never used (the check for end of
2901        line is "zero bytes in current character"), it caused compiler complaints.
2902        I've changed it to 0xffffffff.
2904     8. In pcre_version.c, the version string was being built by a sequence of
2905        C macros that, in the event of PCRE_PRERELEASE being defined as an empty
2906        string (as it is for production releases) called a macro with an empty
2907        argument. The C standard says the result of this is undefined. The gcc
2908        compiler treats it as an empty string (which was what was wanted) but it is
2909        reported that Visual C gives an error. The source has been hacked around to
2910        avoid this problem.
2912     9. On the advice of a Windows user, included <io.h> and <fcntl.h> in Windows
2913        builds of pcretest, and changed the call to _setmode() to use _O_BINARY
2914        instead of 0x8000. Made all the #ifdefs test both _WIN32 and WIN32 (not all
2915        of them did).
2917    10. Originally, pcretest opened its input and output without "b"; then I was
2918        told that "b" was needed in some environments, so it was added for release
2919        5.0 to both the input and output. (It makes no difference on Unix-like
2920        systems.) Later I was told that it is wrong for the input on Windows. I've
2921        now abstracted the modes into two macros, to make it easier to fiddle with
2922        them, and removed "b" from the input mode under Windows.
2924    11. Added pkgconfig support for the C++ wrapper library, libpcrecpp.
2926    12. Added -help and --help to pcretest as an official way of being reminded
2927        of the options.
2929    13. Removed some redundant semicolons after macro calls in pcrecpparg.h.in
2930        and pcrecpp.cc because they annoy compilers at high warning levels.
2932    14. A bit of tidying/refactoring in pcre_exec.c in the main bumpalong loop.
2934    15. Fixed an occurrence of == in configure.ac that should have been = (shell
2935        scripts are not C programs :-) and which was not noticed because it works
2936        on Linux.
2938    16. pcretest is supposed to handle any length of pattern and data line (as one
2939        line or as a continued sequence of lines) by extending its input buffer if
2940        necessary. This feature was broken for very long pattern lines, leading to
2941        a string of junk being passed to pcre_compile() if the pattern was longer
2942        than about 50K.
2944    17. I have done a major re-factoring of the way pcre_compile() computes the
2945        amount of memory needed for a compiled pattern. Previously, there was code
2946        that made a preliminary scan of the pattern in order to do this. That was
2947        OK when PCRE was new, but as the facilities have expanded, it has become
2948        harder and harder to keep it in step with the real compile phase, and there
2949        have been a number of bugs (see for example, 4 above). I have now found a
2950        cunning way of running the real compile function in a "fake" mode that
2951        enables it to compute how much memory it would need, while actually only
2952        ever using a few hundred bytes of working memory and without too many
2953        tests of the mode. This should make future maintenance and development
2954        easier. A side effect of this work is that the limit of 200 on the nesting
2955        depth of parentheses has been removed (though this was never a serious
2956        limitation, I suspect). However, there is a downside: pcre_compile() now
2957        runs more slowly than before (30% or more, depending on the pattern). I
2958        hope this isn't a big issue. There is no effect on runtime performance.
2960    18. Fixed a minor bug in pcretest: if a pattern line was not terminated by a
2961        newline (only possible for the last line of a file) and it was a
2962        pattern that set a locale (followed by /Lsomething), pcretest crashed.
2964    19. Added additional timing features to pcretest. (1) The -tm option now times
2965        matching only, not compiling. (2) Both -t and -tm can be followed, as a
2966        separate command line item, by a number that specifies the number of
2967        repeats to use when timing. The default is 50000; this gives better
2968        precision, but takes uncomfortably long for very large patterns.
2970    20. Extended pcre_study() to be more clever in cases where a branch of a
2971        subpattern has no definite first character. For example, (a*|b*)[cd] would
2972        previously give no result from pcre_study(). Now it recognizes that the
2973        first character must be a, b, c, or d.
2975    21. There was an incorrect error "recursive call could loop indefinitely" if
2976        a subpattern (or the entire pattern) that was being tested for matching an
2977        empty string contained only one non-empty item after a nested subpattern.
2978        For example, the pattern (?>\x{100}*)\d(?R) provoked this error
2979        incorrectly, because the \d was being skipped in the check.
2981    22. The pcretest program now has a new pattern option /B and a command line
2982        option -b, which is equivalent to adding /B to every pattern. This causes
2983        it to show the compiled bytecode, without the additional information that
2984        -d shows. The effect of -d is now the same as -b with -i (and similarly, /D
2985        is the same as /B/I).
2987    23. A new optimization is now able automatically to treat some sequences such
2988        as a*b as a*+b. More specifically, if something simple (such as a character
2989        or a simple class like \d) has an unlimited quantifier, and is followed by
2990        something that cannot possibly match the quantified thing, the quantifier
2991        is automatically "possessified".
2993    24. A recursive reference to a subpattern whose number was greater than 39
2994        went wrong under certain circumstances in UTF-8 mode. This bug could also
2995        have affected the operation of pcre_study().
2997    25. Realized that a little bit of performance could be had by replacing
2998        (c & 0xc0) == 0xc0 with c >= 0xc0 when processing UTF-8 characters.
3000    26. Timing data from pcretest is now shown to 4 decimal places instead of 3.
3002    27. Possessive quantifiers such as a++ were previously implemented by turning
3003        them into atomic groups such as ($>a+). Now they have their own opcodes,
3004        which improves performance. This includes the automatically created ones
3005        from 23 above.
3007    28. A pattern such as (?=(\w+))\1: which simulates an atomic group using a
3008        lookahead was broken if it was not anchored. PCRE was mistakenly expecting
3009        the first matched character to be a colon. This applied both to named and
3010        numbered groups.
3012    29. The ucpinternal.h header file was missing its idempotency #ifdef.
3014    30. I was sent a "project" file called libpcre.a.dev which I understand makes
3015        building PCRE on Windows easier, so I have included it in the distribution.
3017    31. There is now a check in pcretest against a ridiculously large number being
3018        returned by pcre_exec() or pcre_dfa_exec(). If this happens in a /g or /G
3019        loop, the loop is abandoned.
3021    32. Forward references to subpatterns in conditions such as (?(2)...) where
3022        subpattern 2 is defined later cause pcre_compile() to search forwards in
3023        the pattern for the relevant set of parentheses. This search went wrong
3024        when there were unescaped parentheses in a character class, parentheses
3025        escaped with \Q...\E, or parentheses in a #-comment in /x mode.
3027    33. "Subroutine" calls and backreferences were previously restricted to
3028        referencing subpatterns earlier in the regex. This restriction has now
3029        been removed.
3031    34. Added a number of extra features that are going to be in Perl 5.10. On the
3032        whole, these are just syntactic alternatives for features that PCRE had
3033        previously implemented using the Python syntax or my own invention. The
3034        other formats are all retained for compatibility.
3036        (a) Named groups can now be defined as (?<name>...) or (?'name'...) as well
3037            as (?P<name>...). The new forms, as well as being in Perl 5.10, are
3038            also .NET compatible.
3040        (b) A recursion or subroutine call to a named group can now be defined as
3041            (?&name) as well as (?P>name).
3043        (c) A backreference to a named group can now be defined as \k<name> or
3044            \k'name' as well as (?P=name). The new forms, as well as being in Perl
3045            5.10, are also .NET compatible.
3047        (d) A conditional reference to a named group can now use the syntax
3048            (?(<name>) or (?('name') as well as (?(name).
3050        (e) A "conditional group" of the form (?(DEFINE)...) can be used to define
3051            groups (named and numbered) that are never evaluated inline, but can be
3052            called as "subroutines" from elsewhere. In effect, the DEFINE condition
3053            is always false. There may be only one alternative in such a group.
3055        (f) A test for recursion can be given as (?(R1).. or (?(R&name)... as well
3056            as the simple (?(R). The condition is true only if the most recent
3057            recursion is that of the given number or name. It does not search out
3058            through the entire recursion stack.
3060        (g) The escape \gN or \g{N} has been added, where N is a positive or
3061            negative number, specifying an absolute or relative reference.
3063    35. Tidied to get rid of some further signed/unsigned compiler warnings and
3064        some "unreachable code" warnings.
3066    36. Updated the Unicode property tables to Unicode version 5.0.0. Amongst other
3067        things, this adds five new scripts.
3069    37. Perl ignores orphaned \E escapes completely. PCRE now does the same.
3070        There were also incompatibilities regarding the handling of \Q..\E inside
3071        character classes, for example with patterns like [\Qa\E-\Qz\E] where the
3072        hyphen was adjacent to \Q or \E. I hope I've cleared all this up now.
3074    38. Like Perl, PCRE detects when an indefinitely repeated parenthesized group
3075        matches an empty string, and forcibly breaks the loop. There were bugs in
3076        this code in non-simple cases. For a pattern such as  ^(a()*)*  matched
3077        against  aaaa  the result was just "a" rather than "aaaa", for example. Two
3078        separate and independent bugs (that affected different cases) have been
3079        fixed.
3081    39. Refactored the code to abolish the use of different opcodes for small
3082        capturing bracket numbers. This is a tidy that I avoided doing when I
3083        removed the limit on the number of capturing brackets for 3.5 back in 2001.
3084        The new approach is not only tidier, it makes it possible to reduce the
3085        memory needed to fix the previous bug (38).
3087    40. Implemented PCRE_NEWLINE_ANY to recognize any of the Unicode newline
3088        sequences (http://unicode.org/unicode/reports/tr18/) as "newline" when
3089        processing dot, circumflex, or dollar metacharacters, or #-comments in /x
3090        mode.
3092    41. Add \R to match any Unicode newline sequence, as suggested in the Unicode
3093        report.
3095    42. Applied patch, originally from Ari Pollak, modified by Google, to allow
3096        copy construction and assignment in the C++ wrapper.
3098    43. Updated pcregrep to support "--newline=any". In the process, I fixed a
3099        couple of bugs that could have given wrong results in the "--newline=crlf"
3100        case.
3102    44. Added a number of casts and did some reorganization of signed/unsigned int
3103        variables following suggestions from Dair Grant. Also renamed the variable
3104        "this" as "item" because it is a C++ keyword.
3106    45. Arranged for dftables to add
3108          #include "pcre_internal.h"
3110        to pcre_chartables.c because without it, gcc 4.x may remove the array
3111        definition from the final binary if PCRE is built into a static library and
3112        dead code stripping is activated.
3114    46. For an unanchored pattern, if a match attempt fails at the start of a
3115        newline sequence, and the newline setting is CRLF or ANY, and the next two
3116        characters are CRLF, advance by two characters instead of one.
3119    Version 6.7 04-Jul-06
3120    ---------------------
3122     1. In order to handle tests when input lines are enormously long, pcretest has
3123        been re-factored so that it automatically extends its buffers when
3124        necessary. The code is crude, but this _is_ just a test program. The
3125        default size has been increased from 32K to 50K.
3127     2. The code in pcre_study() was using the value of the re argument before
3128        testing it for NULL. (Of course, in any sensible call of the function, it
3129        won't be NULL.)
3131     3. The memmove() emulation function in pcre_internal.h, which is used on
3132        systems that lack both memmove() and bcopy() - that is, hardly ever -
3133        was missing a "static" storage class specifier.
3135     4. When UTF-8 mode was not set, PCRE looped when compiling certain patterns
3136        containing an extended class (one that cannot be represented by a bitmap
3137        because it contains high-valued characters or Unicode property items, e.g.
3138        [\pZ]). Almost always one would set UTF-8 mode when processing such a
3139        pattern, but PCRE should not loop if you do not (it no longer does).
3140        [Detail: two cases were found: (a) a repeated subpattern containing an
3141        extended class; (b) a recursive reference to a subpattern that followed a
3142        previous extended class. It wasn't skipping over the extended class
3143        correctly when UTF-8 mode was not set.]
3145     5. A negated single-character class was not being recognized as fixed-length
3146        in lookbehind assertions such as (?<=[^f]), leading to an incorrect
3147        compile error "lookbehind assertion is not fixed length".
3149     6. The RunPerlTest auxiliary script was showing an unexpected difference
3150        between PCRE and Perl for UTF-8 tests. It turns out that it is hard to
3151        write a Perl script that can interpret lines of an input file either as
3152        byte characters or as UTF-8, which is what "perltest" was being required to
3153        do for the non-UTF-8 and UTF-8 tests, respectively. Essentially what you
3154        can't do is switch easily at run time between having the "use utf8;" pragma
3155        or not. In the end, I fudged it by using the RunPerlTest script to insert
3156        "use utf8;" explicitly for the UTF-8 tests.
3158     7. In multiline (/m) mode, PCRE was matching ^ after a terminating newline at
3159        the end of the subject string, contrary to the documentation and to what
3160        Perl does. This was true of both matching functions. Now it matches only at
3161        the start of the subject and immediately after *internal* newlines.
3163     8. A call of pcre_fullinfo() from pcretest to get the option bits was passing
3164        a pointer to an int instead of a pointer to an unsigned long int. This
3165        caused problems on 64-bit systems.
3167     9. Applied a patch from the folks at Google to pcrecpp.cc, to fix "another
3168        instance of the 'standard' template library not being so standard".
3170    10. There was no check on the number of named subpatterns nor the maximum
3171        length of a subpattern name. The product of these values is used to compute
3172        the size of the memory block for a compiled pattern. By supplying a very
3173        long subpattern name and a large number of named subpatterns, the size
3174        computation could be caused to overflow. This is now prevented by limiting
3175        the length of names to 32 characters, and the number of named subpatterns
3176        to 10,000.
3178    11. Subpatterns that are repeated with specific counts have to be replicated in
3179        the compiled pattern. The size of memory for this was computed from the
3180        length of the subpattern and the repeat count. The latter is limited to
3181        65535, but there was no limit on the former, meaning that integer overflow
3182        could in principle occur. The compiled length of a repeated subpattern is
3183        now limited to 30,000 bytes in order to prevent this.
3185    12. Added the optional facility to have named substrings with the same name.
3187    13. Added the ability to use a named substring as a condition, using the
3188        Python syntax: (?(name)yes|no). This overloads (?(R)... and names that
3189        are numbers (not recommended). Forward references are permitted.
3191    14. Added forward references in named backreferences (if you see what I mean).
3193    15. In UTF-8 mode, with the PCRE_DOTALL option set, a quantified dot in the
3194        pattern could run off the end of the subject. For example, the pattern
3195        "(?s)(.{1,5})"8 did this with the subject "ab".
3197    16. If PCRE_DOTALL or PCRE_MULTILINE were set, pcre_dfa_exec() behaved as if
3198        PCRE_CASELESS was set when matching characters that were quantified with ?
3199        or *.
3201    17. A character class other than a single negated character that had a minimum
3202        but no maximum quantifier - for example [ab]{6,} - was not handled
3203        correctly by pce_dfa_exec(). It would match only one character.
3205    18. A valid (though odd) pattern that looked like a POSIX character
3206        class but used an invalid character after [ (for example [[,abc,]]) caused
3207        pcre_compile() to give the error "Failed: internal error: code overflow" or
3208        in some cases to crash with a glibc free() error. This could even happen if
3209        the pattern terminated after [[ but there just happened to be a sequence of
3210        letters, a binary zero, and a closing ] in the memory that followed.
3212    19. Perl's treatment of octal escapes in the range \400 to \777 has changed
3213        over the years. Originally (before any Unicode support), just the bottom 8
3214        bits were taken. Thus, for example, \500 really meant \100. Nowadays the
3215        output from "man perlunicode" includes this:
3217          The regular expression compiler produces polymorphic opcodes.  That
3218          is, the pattern adapts to the data and automatically switches to
3219          the Unicode character scheme when presented with Unicode data--or
3220          instead uses a traditional byte scheme when presented with byte
3221          data.
3223        Sadly, a wide octal escape does not cause a switch, and in a string with
3224        no other multibyte characters, these octal escapes are treated as before.
3225        Thus, in Perl, the pattern  /\500/ actually matches \100 but the pattern
3226        /\500|\x{1ff}/ matches \500 or \777 because the whole thing is treated as a
3227        Unicode string.
3229        I have not perpetrated such confusion in PCRE. Up till now, it took just
3230        the bottom 8 bits, as in old Perl. I have now made octal escapes with
3231        values greater than \377 illegal in non-UTF-8 mode. In UTF-8 mode they
3232        translate to the appropriate multibyte character.
3234    29. Applied some refactoring to reduce the number of warnings from Microsoft
3235        and Borland compilers. This has included removing the fudge introduced
3236        seven years ago for the OS/2 compiler (see 2.02/2 below) because it caused
3237        a warning about an unused variable.
3239    21. PCRE has not included VT (character 0x0b) in the set of whitespace
3240        characters since release 4.0, because Perl (from release 5.004) does not.
3241        [Or at least, is documented not to: some releases seem to be in conflict
3242        with the documentation.] However, when a pattern was studied with
3243        pcre_study() and all its branches started with \s, PCRE still included VT
3244        as a possible starting character. Of course, this did no harm; it just
3245        caused an unnecessary match attempt.
3247    22. Removed a now-redundant internal flag bit that recorded the fact that case
3248        dependency changed within the pattern. This was once needed for "required
3249        byte" processing, but is no longer used. This recovers a now-scarce options
3250        bit. Also moved the least significant internal flag bit to the most-
3251        significant bit of the word, which was not previously used (hangover from
3252        the days when it was an int rather than a uint) to free up another bit for
3253        the future.
3255    23. Added support for CRLF line endings as well as CR and LF. As well as the
3256        default being selectable at build time, it can now be changed at runtime
3257        via the PCRE_NEWLINE_xxx flags. There are now options for pcregrep to
3258        specify that it is scanning data with non-default line endings.
3260    24. Changed the definition of CXXLINK to make it agree with the definition of
3261        LINK in the Makefile, by replacing LDFLAGS to CXXFLAGS.
3263    25. Applied Ian Taylor's patches to avoid using another stack frame for tail
3264        recursions. This makes a big different to stack usage for some patterns.
3266    26. If a subpattern containing a named recursion or subroutine reference such
3267        as (?P>B) was quantified, for example (xxx(?P>B)){3}, the calculation of
3268        the space required for the compiled pattern went wrong and gave too small a
3269        value. Depending on the environment, this could lead to "Failed: internal
3270        error: code overflow at offset 49" or "glibc detected double free or
3271        corruption" errors.
3273    27. Applied patches from Google (a) to support the new newline modes and (b) to
3274        advance over multibyte UTF-8 characters in GlobalReplace.
3276    28. Change free() to pcre_free() in pcredemo.c. Apparently this makes a
3277        difference for some implementation of PCRE in some Windows version.
3279    29. Added some extra testing facilities to pcretest:
3281        \q<number>   in a data line sets the "match limit" value
3282        \Q<number>   in a data line sets the "match recursion limt" value
3283        -S <number>  sets the stack size, where <number> is in megabytes
3285        The -S option isn't available for Windows.
3288    Version 6.6 06-Feb-06
3289    ---------------------
3291     1. Change 16(a) for 6.5 broke things, because PCRE_DATA_SCOPE was not defined
3292        in pcreposix.h. I have copied the definition from pcre.h.
3294     2. Change 25 for 6.5 broke compilation in a build directory out-of-tree
3295        because pcre.h is no longer a built file.
3297     3. Added Jeff Friedl's additional debugging patches to pcregrep. These are
3298        not normally included in the compiled code.
3301    Version 6.5 01-Feb-06
3302    ---------------------
3304     1. When using the partial match feature with pcre_dfa_exec(), it was not
3305        anchoring the second and subsequent partial matches at the new starting
3306        point. This could lead to incorrect results. For example, with the pattern
3307        /1234/, partially matching against "123" and then "a4" gave a match.
3309     2. Changes to pcregrep:
3311        (a) All non-match returns from pcre_exec() were being treated as failures
3312            to match the line. Now, unless the error is PCRE_ERROR_NOMATCH, an
3313            error message is output. Some extra information is given for the
3314            PCRE_ERROR_MATCHLIMIT and PCRE_ERROR_RECURSIONLIMIT errors, which are
3315            probably the only errors that are likely to be caused by users (by
3316            specifying a regex that has nested indefinite repeats, for instance).
3317            If there are more than 20 of these errors, pcregrep is abandoned.
3319        (b) A binary zero was treated as data while matching, but terminated the
3320            output line if it was written out. This has been fixed: binary zeroes
3321            are now no different to any other data bytes.
3323        (c) Whichever of the LC_ALL or LC_CTYPE environment variables is set is
3324            used to set a locale for matching. The --locale=xxxx long option has
3325            been added (no short equivalent) to specify a locale explicitly on the
3326            pcregrep command, overriding the environment variables.
3328        (d) When -B was used with -n, some line numbers in the output were one less
3329            than they should have been.
3331        (e) Added the -o (--only-matching) option.
3333        (f) If -A or -C was used with -c (count only), some lines of context were
3334            accidentally printed for the final match.
3336        (g) Added the -H (--with-filename) option.
3338        (h) The combination of options -rh failed to suppress file names for files
3339            that were found from directory arguments.
3341        (i) Added the -D (--devices) and -d (--directories) options.
3343        (j) Added the -F (--fixed-strings) option.
3345        (k) Allow "-" to be used as a file name for -f as well as for a data file.
3347        (l) Added the --colo(u)r option.
3349        (m) Added Jeffrey Friedl's -S testing option, but within #ifdefs so that it
3350            is not present by default.
3352     3. A nasty bug was discovered in the handling of recursive patterns, that is,
3353        items such as (?R) or (?1), when the recursion could match a number of
3354        alternatives. If it matched one of the alternatives, but subsequently,
3355        outside the recursion, there was a failure, the code tried to back up into
3356        the recursion. However, because of the way PCRE is implemented, this is not
3357        possible, and the result was an incorrect result from the match.
3359        In order to prevent this happening, the specification of recursion has
3360        been changed so that all such subpatterns are automatically treated as
3361        atomic groups. Thus, for example, (?R) is treated as if it were (?>(?R)).
3363     4. I had overlooked the fact that, in some locales, there are characters for
3364        which isalpha() is true but neither isupper() nor islower() are true. In
3365        the fr_FR locale, for instance, the \xAA and \xBA characters (ordmasculine
3366        and ordfeminine) are like this. This affected the treatment of \w and \W
3367        when they appeared in character classes, but not when they appeared outside
3368        a character class. The bit map for "word" characters is now created
3369        separately from the results of isalnum() instead of just taking it from the
3370        upper, lower, and digit maps. (Plus the underscore character, of course.)
3372     5. The above bug also affected the handling of POSIX character classes such as
3373        [[:alpha:]] and [[:alnum:]]. These do not have their own bit maps in PCRE's
3374        permanent tables. Instead, the bit maps for such a class were previously
3375        created as the appropriate unions of the upper, lower, and digit bitmaps.
3376        Now they are created by subtraction from the [[:word:]] class, which has
3377        its own bitmap.
3379     6. The [[:blank:]] character class matches horizontal, but not vertical space.
3380        It is created by subtracting the vertical space characters (\x09, \x0a,
3381        \x0b, \x0c) from the [[:space:]] bitmap. Previously, however, the
3382        subtraction was done in the overall bitmap for a character class, meaning
3383        that a class such as [\x0c[:blank:]] was incorrect because \x0c would not
3384        be recognized. This bug has been fixed.
3386     7. Patches from the folks at Google:
3388          (a) pcrecpp.cc: "to handle a corner case that may or may not happen in
3389          real life, but is still worth protecting against".
3391          (b) pcrecpp.cc: "corrects a bug when negative radixes are used with
3392          regular expressions".
3394          (c) pcre_scanner.cc: avoid use of std::count() because not all systems
3395          have it.
3397          (d) Split off pcrecpparg.h from pcrecpp.h and had the former built by
3398          "configure" and the latter not, in order to fix a problem somebody had
3399          with compiling the Arg class on HP-UX.
3401          (e) Improve the error-handling of the C++ wrapper a little bit.
3403          (f) New tests for checking recursion limiting.
3405     8. The pcre_memmove() function, which is used only if the environment does not
3406        have a standard memmove() function (and is therefore rarely compiled),
3407        contained two bugs: (a) use of int instead of size_t, and (b) it was not
3408        returning a result (though PCRE never actually uses the result).
3410     9. In the POSIX regexec() interface, if nmatch is specified as a ridiculously
3411        large number - greater than INT_MAX/(3*sizeof(int)) - REG_ESPACE is
3412        returned instead of calling malloc() with an overflowing number that would
3413        most likely cause subsequent chaos.
3415    10. The debugging option of pcretest was not showing the NO_AUTO_CAPTURE flag.
3417    11. The POSIX flag REG_NOSUB is now supported. When a pattern that was compiled
3418        with this option is matched, the nmatch and pmatch options of regexec() are
3419        ignored.
3421    12. Added REG_UTF8 to the POSIX interface. This is not defined by POSIX, but is
3422        provided in case anyone wants to the the POSIX interface with UTF-8
3423        strings.
3425    13. Added CXXLDFLAGS to the Makefile parameters to provide settings only on the
3426        C++ linking (needed for some HP-UX environments).
3428    14. Avoid compiler warnings in get_ucpname() when compiled without UCP support
3429        (unused parameter) and in the pcre_printint() function (omitted "default"
3430        switch label when the default is to do nothing).
3432    15. Added some code to make it possible, when PCRE is compiled as a C++
3433        library, to replace subject pointers for pcre_exec() with a smart pointer
3434        class, thus making it possible to process discontinuous strings.
3436    16. The two macros PCRE_EXPORT and PCRE_DATA_SCOPE are confusing, and perform
3437        much the same function. They were added by different people who were trying
3438        to make PCRE easy to compile on non-Unix systems. It has been suggested
3439        that PCRE_EXPORT be abolished now that there is more automatic apparatus
3440        for compiling on Windows systems. I have therefore replaced it with
3441        PCRE_DATA_SCOPE. This is set automatically for Windows; if not set it
3442        defaults to "extern" for C or "extern C" for C++, which works fine on
3443        Unix-like systems. It is now possible to override the value of PCRE_DATA_
3444        SCOPE with something explicit in config.h. In addition:
3446        (a) pcreposix.h still had just "extern" instead of either of these macros;
3447            I have replaced it with PCRE_DATA_SCOPE.
3449        (b) Functions such as _pcre_xclass(), which are internal to the library,
3450            but external in the C sense, all had PCRE_EXPORT in their definitions.
3451            This is apparently wrong for the Windows case, so I have removed it.
3452            (It makes no difference on Unix-like systems.)
3454    17. Added a new limit, MATCH_LIMIT_RECURSION, which limits the depth of nesting
3455        of recursive calls to match(). This is different to MATCH_LIMIT because
3456        that limits the total number of calls to match(), not all of which increase
3457        the depth of recursion. Limiting the recursion depth limits the amount of
3458        stack (or heap if NO_RECURSE is set) that is used. The default can be set
3459        when PCRE is compiled, and changed at run time. A patch from Google adds
3460        this functionality to the C++ interface.
3462    18. Changes to the handling of Unicode character properties:
3464        (a) Updated the table to Unicode 4.1.0.
3466        (b) Recognize characters that are not in the table as "Cn" (undefined).
3468        (c) I revised the way the table is implemented to a much improved format
3469            which includes recognition of ranges. It now supports the ranges that
3470            are defined in UnicodeData.txt, and it also amalgamates other
3471            characters into ranges. This has reduced the number of entries in the
3472            table from around 16,000 to around 3,000, thus reducing its size
3473            considerably. I realized I did not need to use a tree structure after
3474            all - a binary chop search is just as efficient. Having reduced the
3475            number of entries, I extended their size from 6 bytes to 8 bytes to
3476            allow for more data.
3478        (d) Added support for Unicode script names via properties such as \p{Han}.
3480    19. In UTF-8 mode, a backslash followed by a non-Ascii character was not
3481        matching that character.
3483    20. When matching a repeated Unicode property with a minimum greater than zero,
3484        (for example \pL{2,}), PCRE could look past the end of the subject if it
3485        reached it while seeking the minimum number of characters. This could
3486        happen only if some of the characters were more than one byte long, because
3487        there is a check for at least the minimum number of bytes.
3489    21. Refactored the implementation of \p and \P so as to be more general, to
3490        allow for more different types of property in future. This has changed the
3491        compiled form incompatibly. Anybody with saved compiled patterns that use
3492        \p or \P will have to recompile them.
3494    22. Added "Any" and "L&" to the supported property types.
3496    23. Recognize \x{...} as a code point specifier, even when not in UTF-8 mode,
3497        but give a compile time error if the value is greater than 0xff.
3499    24. The man pages for pcrepartial, pcreprecompile, and pcre_compile2 were
3500        accidentally not being installed or uninstalled.
3502    25. The pcre.h file was built from pcre.h.in, but the only changes that were
3503        made were to insert the current release number. This seemed silly, because
3504        it made things harder for people building PCRE on systems that don't run
3505        "configure". I have turned pcre.h into a distributed file, no longer built
3506        by "configure", with the version identification directly included. There is
3507        no longer a pcre.h.in file.
3509        However, this change necessitated a change to the pcre-config script as
3510        well. It is built from pcre-config.in, and one of the substitutions was the
3511        release number. I have updated configure.ac so that ./configure now finds
3512        the release number by grepping pcre.h.
3514    26. Added the ability to run the tests under valgrind.
3517    Version 6.4 05-Sep-05
3518    ---------------------
3520     1. Change 6.0/10/(l) to pcregrep introduced a bug that caused separator lines
3521        "--" to be printed when multiple files were scanned, even when none of the
3522        -A, -B, or -C options were used. This is not compatible with Gnu grep, so I
3523        consider it to be a bug, and have restored the previous behaviour.
3525     2. A couple of code tidies to get rid of compiler warnings.
3527     3. The pcretest program used to cheat by referring to symbols in the library
3528        whose names begin with _pcre_. These are internal symbols that are not
3529        really supposed to be visible externally, and in some environments it is
3530        possible to suppress them. The cheating is now confined to including
3531        certain files from the library's source, which is a bit cleaner.
3533     4. Renamed pcre.in as pcre.h.in to go with pcrecpp.h.in; it also makes the
3534        file's purpose clearer.
3536     5. Reorganized pcre_ucp_findchar().
3539    Version 6.3 15-Aug-05
3540    ---------------------
3542     1. The file libpcre.pc.in did not have general read permission in the tarball.
3544     2. There were some problems when building without C++ support:
3546        (a) If C++ support was not built, "make install" and "make test" still
3547            tried to test it.
3549        (b) There were problems when the value of CXX was explicitly set. Some
3550            changes have been made to try to fix these, and ...
3552        (c) --disable-cpp can now be used to explicitly disable C++ support.
3554        (d) The use of @CPP_OBJ@ directly caused a blank line preceded by a
3555            backslash in a target when C++ was disabled. This confuses some
3556            versions of "make", apparently. Using an intermediate variable solves
3557            this. (Same for CPP_LOBJ.)
3559     3. $(LINK_FOR_BUILD) now includes $(CFLAGS_FOR_BUILD) and $(LINK)
3560        (non-Windows) now includes $(CFLAGS) because these flags are sometimes
3561        necessary on certain architectures.
3563     4. Added a setting of -export-symbols-regex to the link command to remove
3564        those symbols that are exported in the C sense, but actually are local
3565        within the library, and not documented. Their names all begin with
3566        "_pcre_". This is not a perfect job, because (a) we have to except some
3567        symbols that pcretest ("illegally") uses, and (b) the facility isn't always
3568        available (and never for static libraries). I have made a note to try to
3569        find a way round (a) in the future.
3572    Version 6.2 01-Aug-05
3573    ---------------------
3575     1. There was no test for integer overflow of quantifier values. A construction
3576        such as {1111111111111111} would give undefined results. What is worse, if
3577        a minimum quantifier for a parenthesized subpattern overflowed and became
3578        negative, the calculation of the memory size went wrong. This could have
3579        led to memory overwriting.
3581     2. Building PCRE using VPATH was broken. Hopefully it is now fixed.