Merge branch 'rb/stdlib_re_unicode_fixes' into ccase/r13b04_dev

* rb/stdlib_re_unicode_fixes:
  Fix lost unicode option in re:compile()
  Refactor out repeated block in re module
  Fix re:replace/4 to handle unicode charlist Replacement argument
  Fix re:replace/4 to handle unicode charlist RE argument
  Fix re:replace/4 to handle binary unicode output when nothing replaced

OTP-8394  A number of bugs concerning re and unicode are corrected:
          - re:compile no longer loses unicode option, which also fixes bug
            in re:split.
          - re:replace now handles unicode charlist replacement argument
          - re:replace now handles unicode RE charlist argument correctly
          - re:replace now handles binary unicode output correctly when
          nothing is replaced.

          Most code, testcases and error isolation done by Rory Byrne.

5 files changed, 957 insertions, 73 deletions

diff --git a/erts/emulator/beam/erl_bif_re.c b/erts/emulator/beam/erl_bif_re.c
index 6efc19597b..c027cd5984 100644
--- a/erts/emulator/beam/erl_bif_re.c
+++ b/erts/emulator/beam/erl_bif_re.c
@@ -884,7 +884,7 @@ re_run_3(BIF_ALIST_3)
 	    int capture_count;
 
 	    if (pflags & PARSE_FLAG_UNICODE && 
-		(!is_binary(BIF_ARG_1) || 
+		(!is_binary(BIF_ARG_2) || !is_binary(BIF_ARG_1) ||
 		 (is_list_cap && !(pflags & PARSE_FLAG_GLOBAL)))) { 
 		BIF_TRAP3(urun_trap_exportp, BIF_P, BIF_ARG_1, BIF_ARG_2, BIF_ARG_3);
 	    }
diff --git a/lib/stdlib/src/re.erl b/lib/stdlib/src/re.erl
index 5417ac02e5..296a6b3d23 100644
--- a/lib/stdlib/src/re.erl
+++ b/lib/stdlib/src/re.erl
@@ -1,19 +1,19 @@
 %%
 %% %CopyrightBegin%
-%% 
-%% Copyright Ericsson AB 2008-2009. All Rights Reserved.
-%% 
+%%
+%% Copyright Ericsson AB 2008-2010. All Rights Reserved.
+%%
 %% The contents of this file are subject to the Erlang Public License,
 %% Version 1.1, (the "License"); you may not use this file except in
 %% compliance with the License. You should have received a copy of the
 %% Erlang Public License along with this software. If not, it can be
 %% retrieved online at http://www.erlang.org/.
-%% 
+%%
 %% Software distributed under the License is distributed on an "AS IS"
 %% basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
 %% the License for the specific language governing rights and limitations
 %% under the License.
-%% 
+%%
 %% %CopyrightEnd%
 %%
 -module(re).
@@ -32,18 +32,7 @@ split(Subject,RE,Options) ->
     try
     {NewOpt,Convert,Unicode,Limit,Strip,Group} =
 	process_split_params(Options,iodata,false,-1,false,false),
-    FlatSubject = 
-	case is_binary(Subject) of
-	    true ->
-		Subject;
-	    false ->
-		case Unicode of
-		    true ->
-			unicode:characters_to_binary(Subject,unicode);
-		    false ->
-			iolist_to_binary(Subject)
-		end
-	end,
+    FlatSubject = to_binary(Subject, Unicode),
     case compile_split(RE,NewOpt) of
 	{error,_Err} ->
 	    throw(badre);
@@ -217,19 +206,9 @@ replace(Subject,RE,Replacement,Options) ->
     try
     {NewOpt,Convert,Unicode} =
 	process_repl_params(Options,iodata,false),
-    FlatSubject = 
-	case is_binary(Subject) of
-	    true ->
-		Subject;
-	    false ->
-		case Unicode of
-		    true ->
-			unicode:characters_to_binary(Subject,unicode);
-		    false ->
-			iolist_to_binary(Subject)
-		end
-	end,
-    case do_replace(FlatSubject,Subject,RE,Replacement,NewOpt) of
+    FlatSubject = to_binary(Subject, Unicode),
+    FlatReplacement = to_binary(Replacement, Unicode),
+    case do_replace(FlatSubject,Subject,RE,FlatReplacement,NewOpt) of
 	{error,_Err} ->
 	    throw(badre);
 	IoList ->
@@ -237,7 +216,12 @@ replace(Subject,RE,Replacement,Options) ->
 		iodata ->
 		    IoList;
 		binary ->
-		    iolist_to_binary(IoList);
+		    case Unicode of
+			false ->
+			    iolist_to_binary(IoList);
+			true ->
+			    unicode:characters_to_binary(IoList,unicode)
+		    end;
 		list ->
 		    case Unicode of
 			false ->
@@ -324,8 +308,7 @@ process_split_params([H|T],C,U,L,S,G) ->
     {[H|NT],NC,NU,NL,NS,NG}.
 
 apply_mlist(Subject,Replacement,Mlist) ->
-    do_mlist(Subject,Subject,0,precomp_repl(iolist_to_binary(Replacement)),
-	     Mlist).
+    do_mlist(Subject,Subject,0,precomp_repl(Replacement), Mlist).
 
 
 precomp_repl(<<>>) ->
@@ -545,7 +528,7 @@ process_uparams([],Type) ->
 
 ucompile(RE,Options) ->
     try
-	re:compile(unicode:characters_to_binary(RE,unicode))
+	re:compile(unicode:characters_to_binary(RE,unicode),Options)
     catch
 	error:AnyError ->
 	    {'EXIT',{new_stacktrace,[{Mod,_,L}|Rest]}} = 
@@ -618,18 +601,7 @@ grun(Subject,RE,{Options,NeedClean,OrigRE}) ->
 
 grun2(Subject,RE,{Options,NeedClean}) ->
     Unicode = check_for_unicode(RE,Options),
-    FlatSubject = 
-	case is_binary(Subject) of
-	    true ->
-		Subject;
-	    false ->
-		case Unicode of
-		    true ->
-			unicode:characters_to_binary(Subject,unicode);
-		    false ->
-			iolist_to_binary(Subject)
-		end
-	end,
+    FlatSubject = to_binary(Subject, Unicode),
     do_grun(FlatSubject,Subject,Unicode,RE,{Options,NeedClean}).
 
 do_grun(FlatSubject,Subject,Unicode,RE,{Options0,NeedClean}) ->
@@ -749,3 +721,10 @@ runopt(global) ->
     true;
 runopt(_) ->
     false.
+
+to_binary(Bin, _IsUnicode) when is_binary(Bin) ->
+    Bin;
+to_binary(Data, true) ->
+    unicode:characters_to_binary(Data,unicode);
+to_binary(Data, false) ->
+    iolist_to_binary(Data).
diff --git a/lib/stdlib/test/re_SUITE.erl b/lib/stdlib/test/re_SUITE.erl
index 98eb66d1fb..fa50ba3b7a 100644
--- a/lib/stdlib/test/re_SUITE.erl
+++ b/lib/stdlib/test/re_SUITE.erl
@@ -1,29 +1,29 @@
 %%
 %% %CopyrightBegin%
-%% 
-%% Copyright Ericsson AB 2008-2009. All Rights Reserved.
-%% 
+%%
+%% Copyright Ericsson AB 2008-2010. All Rights Reserved.
+%%
 %% The contents of this file are subject to the Erlang Public License,
 %% Version 1.1, (the "License"); you may not use this file except in
 %% compliance with the License. You should have received a copy of the
 %% Erlang Public License along with this software. If not, it can be
 %% retrieved online at http://www.erlang.org/.
-%% 
+%%
 %% Software distributed under the License is distributed on an "AS IS"
 %% basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
 %% the License for the specific language governing rights and limitations
 %% under the License.
-%% 
+%%
 %% %CopyrightEnd%
 %%
 -module(re_SUITE).
 
--export([all/1, pcre/1,compile_options/1,run_options/1,combined_options/1,replace_autogen/1,global_capture/1,replace_return/1,split_autogen/1,split_options/1,split_specials/1,error_handling/1]).
+-export([all/1, pcre/1,compile_options/1,run_options/1,combined_options/1,replace_autogen/1,global_capture/1,replace_input_types/1,replace_return/1,split_autogen/1,split_options/1,split_specials/1,error_handling/1]).
 
 -include("test_server.hrl").
 -include_lib("kernel/include/file.hrl").
 
-all(suite) -> [pcre,compile_options,run_options,combined_options,replace_autogen,global_capture,replace_return,split_autogen,split_options,split_specials,error_handling].
+all(suite) -> [pcre,compile_options,run_options,combined_options,replace_autogen,global_capture,replace_input_types,replace_return,split_autogen,split_options,split_specials,error_handling].
 
 pcre(doc) ->
     ["Run all applicable tests from the PCRE testsuites."];
@@ -268,7 +268,17 @@ global_capture(Config) when is_list(Config) ->
     ?line {match,[[{3,5},{5,3}],[{11,4},{12,3}]]} = re:run("ABC�bcdABCabcdA",".(?<FOO>bcd)",[global,{capture,all,index},unicode]),
     ?t:timetrap_cancel(Dog),
     ok.
-    
+
+replace_input_types(doc) ->
+    ["Tests replace with different input types"];
+replace_input_types(Config) when is_list(Config) ->
+    Dog = ?t:timetrap(?t:minutes(3)),
+    ?line <<"abcd">> = re:replace("abcd","Z","X",[{return,binary},unicode]),
+    ?line <<"abcd">> = re:replace("abcd","\x{400}","X",[{return,binary},unicode]),
+    ?line <<"a",208,128,"cd">> = re:replace(<<"abcd">>,"b","\x{400}",[{return,binary},unicode]),
+    ?t:timetrap_cancel(Dog),
+    ok.
+
 replace_return(doc) ->    
     ["Tests return options of replace together with global searching"];
 replace_return(Config) when is_list(Config) ->
@@ -289,6 +299,10 @@ replace_return(Config) when is_list(Config) ->
     ?line <<"iXk">> = re:replace("abcdefghijk","(.)(.)(.)(.)(.)(.)(.)(.)(.)(.)","\\9X",[{return,binary}]),
     ?line <<"jXk">> = re:replace("abcdefghijk","(.)(.)(.)(.)(.)(.)(.)(.)(.)(.)","\\10X",[{return,binary}]),
     ?line <<"Xk">> = re:replace("abcdefghijk","(.)(.)(.)(.)(.)(.)(.)(.)(.)(.)","\\11X",[{return,binary}]),
+    ?line "a\x{400}bcX" = re:replace("a\x{400}bcd","d","X",[global,{return,list},unicode]),
+    ?line <<"a",208,128,"bcX">> = re:replace("a\x{400}bcd","d","X",[global,{return,binary},unicode]),
+    ?line "a\x{400}bcd" = re:replace("a\x{400}bcd","Z","X",[global,{return,list},unicode]),
+    ?line <<"a",208,128,"bcd">> = re:replace("a\x{400}bcd","Z","X",[global,{return,binary},unicode]),
     ?t:timetrap_cancel(Dog),
     ok.
 
diff --git a/lib/stdlib/test/re_SUITE_data/mod_testoutput8 b/lib/stdlib/test/re_SUITE_data/mod_testoutput8
new file mode 100644
index 0000000000..b7e7b02d6c
--- /dev/null
+++ b/lib/stdlib/test/re_SUITE_data/mod_testoutput8
@@ -0,0 +1,877 @@
+/-- Do not use the \x{} construct except with patterns that have the --/
+/-- /8 option set, because PCRE doesn't recognize them as UTF-8 unless --/
+No match
+/-- that option is set. However, the latest Perls recognize them always. --/
+No match
+
+\x{100}ab/8
+  \x{100}ab
+ 0: \x{100}ab
+  
+/a\x{100}*b/8
+    ab
+ 0: ab
+    a\x{100}b  
+ 0: a\x{100}b
+    a\x{100}\x{100}b  
+ 0: a\x{100}\x{100}b
+    
+/a\x{100}+b/8
+    a\x{100}b  
+ 0: a\x{100}b
+    a\x{100}\x{100}b  
+ 0: a\x{100}\x{100}b
+    *** Failers 
+No match
+    ab
+No match
+     
+/\bX/8
+    Xoanon
+ 0: X
+    +Xoanon
+ 0: X
+    \x{300}Xoanon 
+ 0: X
+    *** Failers 
+No match
+    YXoanon  
+No match
+    
+/\BX/8
+    YXoanon
+ 0: X
+    *** Failers
+No match
+    Xoanon
+No match
+    +Xoanon    
+No match
+    \x{300}Xoanon 
+No match
+
+/X\b/8
+    X+oanon
+ 0: X
+    ZX\x{300}oanon 
+ 0: X
+    FAX 
+ 0: X
+    *** Failers 
+No match
+    Xoanon  
+No match
+    
+/X\B/8
+    Xoanon  
+ 0: X
+    *** Failers
+No match
+    X+oanon
+No match
+    ZX\x{300}oanon 
+No match
+    FAX 
+No match
+    
+/[^a]/8
+    abcd
+ 0: b
+    a\x{100}   
+ 0: \x{100}
+
+/^[abc\x{123}\x{400}-\x{402}]{2,3}\d/8
+    ab99
+ 0: ab9
+    \x{123}\x{123}45
+ 0: \x{123}\x{123}4
+    \x{400}\x{401}\x{402}6  
+ 0: \x{400}\x{401}\x{402}6
+    *** Failers
+No match
+    d99
+No match
+    \x{123}\x{122}4   
+No match
+    \x{400}\x{403}6  
+No match
+    \x{400}\x{401}\x{402}\x{402}6  
+No match
+
+/abc/8
+    �]
+Error -10
+    �
+Error -10
+    ���
+Error -10
+    ���\?
+No match
+
+/a.b/8
+    acb
+ 0: acb
+    a\x7fb
+ 0: a\x{7f}b
+    a\x{100}b 
+ 0: a\x{100}b
+    *** Failers
+No match
+    a\nb  
+No match
+
+/^[a\x{c0}]/8
+    *** Failers
+No match
+    \x{100}
+No match
+
+/(?<=aXb)cd/8
+    aXbcd
+ 0: cd
+
+/(?<=a\x{100}b)cd/8
+    a\x{100}bcd
+ 0: cd
+
+/(?<=a\x{100000}b)cd/8
+    a\x{100000}bcd
+ 0: cd
+    
+/(?:\x{100}){3}b/8
+    \x{100}\x{100}\x{100}b
+ 0: \x{100}\x{100}\x{100}b
+    *** Failers 
+No match
+    \x{100}\x{100}b
+No match
+
+/\x{ab}/8
+    \x{ab} 
+ 0: \x{ab}
+    \xc2\xab
+ 0: \x{ab}
+    *** Failers 
+No match
+    \x00{ab}
+No match
+
+/^[^a]{2}/8
+    \x{100}bc
+ 0: \x{100}b
+ 
+/^[^a]{2,}/8
+    \x{100}bcAa
+ 0: \x{100}bcA
+
+/^[^a]{2,}?/8
+    \x{100}bca
+ 0: \x{100}b
+
+/^[^a]{2}/8i
+    \x{100}bc
+ 0: \x{100}b
+ 
+/^[^a]{2,}/8i
+    \x{100}bcAa
+ 0: \x{100}bc
+
+/^[^a]{2,}?/8iU
+    \x{100}bca
+ 0: \x{100}bc
+
+/\x{100}{0,0}/8
+    abcd
+ 0: 
+ 
+/\x{100}?/8
+    abcd
+ 0: 
+    \x{100}\x{100} 
+ 0: \x{100}
+
+/\x{100}{0,3}/8 
+    \x{100}\x{100} 
+ 0: \x{100}\x{100}
+    \x{100}\x{100}\x{100}\x{100} 
+ 0: \x{100}\x{100}\x{100}
+    
+/\x{100}*/8
+    abce
+ 0: 
+    \x{100}\x{100}\x{100}\x{100} 
+ 0: \x{100}\x{100}\x{100}\x{100}
+
+/\x{100}{1,1}/8
+    abcd\x{100}\x{100}\x{100}\x{100} 
+ 0: \x{100}
+
+/\x{100}{1,3}/8
+    abcd\x{100}\x{100}\x{100}\x{100} 
+ 0: \x{100}\x{100}\x{100}
+
+/\x{100}+/8
+    abcd\x{100}\x{100}\x{100}\x{100} 
+ 0: \x{100}\x{100}\x{100}\x{100}
+
+/\x{100}{3}/8
+    abcd\x{100}\x{100}\x{100}XX
+ 0: \x{100}\x{100}\x{100}
+
+/\x{100}{3,5}/8
+    abcd\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}XX
+ 0: \x{100}\x{100}\x{100}\x{100}\x{100}
+
+/\x{100}{3,}/8
+    abcd\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}XX
+ 0: \x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}
+
+/(?<=a\x{100}{2}b)X/8
+    Xyyya\x{100}\x{100}bXzzz
+ 0: X
+
+/\D*/8
+  aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+ 0: aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
+
+/\D*/8
+  \x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}
+ 0: \x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}\x{100}
+
+/\D/8
+    1X2
+ 0: X
+    1\x{100}2 
+ 0: \x{100}
+  
+/>\S/8
+    > >X Y
+ 0: >X
+    > >\x{100} Y
+ 0: >\x{100}
+  
+/\d/8
+    \x{100}3
+ 0: 3
+    
+/\s/8
+    \x{100} X
+ 0:  
+    
+/\D+/8
+    12abcd34
+ 0: abcd
+    *** Failers
+ 0: *** Failers
+    1234  
+No match
+
+/\d+/8
+    12abcd34
+ 0: 12
+     *** Failers
+No match
+
+/\d{2,3}/8
+    12abcd34
+ 0: 12
+    1234abcd
+ 0: 123
+     *** Failers  
+No match
+    1.4 
+No match
+
+/\S+/8
+    12abcd34
+ 0: 12abcd34
+    *** Failers
+ 0: ***
+     \    \ 
+No match
+
+/>\s+</8
+    12>      <34
+ 0: >      <
+    *** Failers
+No match
+
+/>\s{2,3}</8
+    ab>  <cd
+ 0: >  <
+    ab>   <ce
+ 0: >   <
+    *** Failers
+No match
+    ab>    <cd 
+No match
+
+/>\s{2,3}?</8
+    ab>  <cd
+ 0: >  <
+    ab>   <ce
+ 0: >   <
+    *** Failers
+No match
+    ab>    <cd 
+No match
+
+/\w+/8
+    12      34
+ 0: 12
+     *** Failers
+ 0: Failers
+    +++=*! 
+No match
+
+/\w{2,3}/8
+    ab  cd
+ 0: ab
+    abcd ce
+ 0: abc
+    *** Failers
+ 0: Fai
+    a.b.c
+No match
+
+/\W+/8
+    12====34
+ 0: ====
+    *** Failers
+ 0: *** 
+    abcd 
+No match
+
+/\W{2,3}/8
+    ab====cd
+ 0: ===
+    ab==cd
+ 0: ==
+    *** Failers
+ 0: ***
+    a.b.c
+No match
+
+/\W{2,3}?/8U
+    ab====cd
+ 0: ===
+     ab==cd
+ 0: ==
+    *** Failers
+ 0: ***
+     a.b.c
+No match
+
+/[\x{100}]/8
+    \x{100}
+ 0: \x{100}
+    Z\x{100}
+ 0: \x{100}
+    \x{100}Z
+ 0: \x{100}
+    *** Failers 
+No match
+
+/[Z\x{100}]/8
+    Z\x{100}
+ 0: Z
+    \x{100}
+ 0: \x{100}
+    \x{100}Z
+ 0: \x{100}
+    *** Failers 
+No match
+
+/[\x{100}\x{200}]/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   *** Failers  
+No match
+
+/[\x{100}-\x{200}]/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{111}cd 
+ 0: \x{111}
+   *** Failers  
+No match
+
+/[z-\x{200}]/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{111}cd 
+ 0: \x{111}
+   abzcd
+ 0: z
+   ab|cd  
+ 0: |
+   *** Failers  
+No match
+
+/[Q\x{100}\x{200}]/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   Q? 
+ 0: Q
+   *** Failers  
+No match
+
+/[Q\x{100}-\x{200}]/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{111}cd 
+ 0: \x{111}
+   Q? 
+ 0: Q
+   *** Failers  
+No match
+
+/[Qz-\x{200}]/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{111}cd 
+ 0: \x{111}
+   abzcd
+ 0: z
+   ab|cd  
+ 0: |
+   Q? 
+ 0: Q
+   *** Failers  
+No match
+
+/[\x{100}\x{200}]{1,3}/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{200}\x{100}\x{200}\x{100}cd
+ 0: \x{200}\x{100}\x{200}
+   *** Failers  
+No match
+
+/[\x{100}\x{200}]{1,3}?/8U
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{200}\x{100}\x{200}\x{100}cd
+ 0: \x{200}\x{100}\x{200}
+   *** Failers  
+No match
+
+/[Q\x{100}\x{200}]{1,3}/8
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{200}\x{100}\x{200}\x{100}cd
+ 0: \x{200}\x{100}\x{200}
+   *** Failers  
+No match
+
+/[Q\x{100}\x{200}]{1,3}?/8U
+   ab\x{100}cd
+ 0: \x{100}
+   ab\x{200}cd
+ 0: \x{200}
+   ab\x{200}\x{100}\x{200}\x{100}cd
+ 0: \x{200}\x{100}\x{200}
+   *** Failers  
+No match
+
+/(?<=[\x{100}\x{200}])X/8
+    abc\x{200}X
+ 0: X
+    abc\x{100}X 
+ 0: X
+    *** Failers
+No match
+    X  
+No match
+
+/(?<=[Q\x{100}\x{200}])X/8
+    abc\x{200}X
+ 0: X
+    abc\x{100}X 
+ 0: X
+    abQX 
+ 0: X
+    *** Failers
+No match
+    X  
+No match
+
+/(?<=[\x{100}\x{200}]{3})X/8
+    abc\x{100}\x{200}\x{100}X
+ 0: X
+    *** Failers
+No match
+    abc\x{200}X
+No match
+    X  
+No match
+
+/[^\x{100}\x{200}]X/8
+    AX
+ 0: AX
+    \x{150}X
+ 0: \x{150}X
+    \x{500}X 
+ 0: \x{500}X
+    *** Failers
+No match
+    \x{100}X
+No match
+    \x{200}X   
+No match
+
+/[^Q\x{100}\x{200}]X/8
+    AX
+ 0: AX
+    \x{150}X
+ 0: \x{150}X
+    \x{500}X 
+ 0: \x{500}X
+    *** Failers
+No match
+    \x{100}X
+No match
+    \x{200}X   
+No match
+    QX 
+No match
+
+/[^\x{100}-\x{200}]X/8
+    AX
+ 0: AX
+    \x{500}X 
+ 0: \x{500}X
+    *** Failers
+No match
+    \x{100}X
+No match
+    \x{150}X
+No match
+    \x{200}X   
+No match
+
+/[z-\x{100}]/8i
+    z
+ 0: z
+    Z 
+ 0: Z
+    \x{100}
+ 0: \x{100}
+    *** Failers
+No match
+    \x{102}
+No match
+    y    
+No match
+
+/[\xFF]/
+    >\xff<
+ 0: \xff
+
+/[\xff]/8
+    >\x{ff}<
+ 0: \x{ff}
+
+/[^\xFF]/
+    XYZ
+ 0: X
+
+/[^\xff]/8
+    XYZ
+ 0: X
+    \x{123} 
+ 0: \x{123}
+
+/^[ac]*b/8
+  xb
+No match
+
+/^[ac\x{100}]*b/8
+  xb
+No match
+
+/^[^x]*b/8i
+  xb
+No match
+
+/^[^x]*b/8
+  xb
+No match
+  
+/^\d*b/8
+  xb 
+No match
+
+/^\x{85}$/8i
+    \x{85}
+ 0: \x{85}
+
+/^abc./mgx8<any>
+    abc1 \x0aabc2 \x0babc3xx \x0cabc4 \x0dabc5xx \x0d\x0aabc6 \x{0085}abc7 \x{2028}abc8 \x{2029}abc9 JUNK
+ 0: abc1
+ 0: abc2
+ 0: abc3
+ 0: abc4
+ 0: abc5
+ 0: abc6
+ 0: abc7
+ 0: abc8
+ 0: abc9
+
+/abc.$/mgx8<any>
+    abc1\x0a abc2\x0b abc3\x0c abc4\x0d abc5\x0d\x0a abc6\x{0085} abc7\x{2028} abc8\x{2029} abc9
+ 0: abc1
+ 0: abc2
+ 0: abc3
+ 0: abc4
+ 0: abc5
+ 0: abc6
+ 0: abc7
+ 0: abc8
+ 0: abc9
+
+/^a\Rb/8<bsr_unicode>
+    a\nb
+ 0: a\x{0a}b
+    a\rb
+ 0: a\x{0d}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    a\x0bb
+ 0: a\x{0b}b
+    a\x0cb
+ 0: a\x{0c}b
+    a\x{85}b   
+ 0: a\x{85}b
+    a\x{2028}b 
+ 0: a\x{2028}b
+    a\x{2029}b 
+ 0: a\x{2029}b
+    ** Failers
+No match
+    a\n\rb    
+No match
+
+/^a\R*b/8<bsr_unicode>
+    ab
+ 0: ab
+    a\nb
+ 0: a\x{0a}b
+    a\rb
+ 0: a\x{0d}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    a\x0bb
+ 0: a\x{0b}b
+    a\x0c\x{2028}\x{2029}b
+ 0: a\x{0c}\x{2028}\x{2029}b
+    a\x{85}b   
+ 0: a\x{85}b
+    a\n\rb    
+ 0: a\x{0a}\x{0d}b
+    a\n\r\x{85}\x0cb 
+ 0: a\x{0a}\x{0d}\x{85}\x{0c}b
+
+/^a\R+b/8<bsr_unicode>
+    a\nb
+ 0: a\x{0a}b
+    a\rb
+ 0: a\x{0d}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    a\x0bb
+ 0: a\x{0b}b
+    a\x0c\x{2028}\x{2029}b
+ 0: a\x{0c}\x{2028}\x{2029}b
+    a\x{85}b   
+ 0: a\x{85}b
+    a\n\rb    
+ 0: a\x{0a}\x{0d}b
+    a\n\r\x{85}\x0cb 
+ 0: a\x{0a}\x{0d}\x{85}\x{0c}b
+    ** Failers
+No match
+    ab  
+No match
+
+/^a\R{1,3}b/8<bsr_unicode>
+    a\nb
+ 0: a\x{0a}b
+    a\n\rb
+ 0: a\x{0a}\x{0d}b
+    a\n\r\x{85}b
+ 0: a\x{0a}\x{0d}\x{85}b
+    a\r\n\r\nb 
+ 0: a\x{0d}\x{0a}\x{0d}\x{0a}b
+    a\r\n\r\n\r\nb 
+ 0: a\x{0d}\x{0a}\x{0d}\x{0a}\x{0d}\x{0a}b
+    a\n\r\n\rb
+ 0: a\x{0a}\x{0d}\x{0a}\x{0d}b
+    a\n\n\r\nb 
+ 0: a\x{0a}\x{0a}\x{0d}\x{0a}b
+    ** Failers
+No match
+    a\n\n\n\rb
+No match
+    a\r
+No match
+
+/\h+\V?\v{3,4}/8 
+    \x09\x20\x{a0}X\x0a\x0b\x0c\x0d\x0a
+ 0: \x{09} \x{a0}X\x{0a}\x{0b}\x{0c}\x{0d}
+
+/\V?\v{3,4}/8 
+    \x20\x{a0}X\x0a\x0b\x0c\x0d\x0a
+ 0: X\x{0a}\x{0b}\x{0c}\x{0d}
+
+/\h+\V?\v{3,4}/8
+    >\x09\x20\x{a0}X\x0a\x0a\x0a<
+ 0: \x{09} \x{a0}X\x{0a}\x{0a}\x{0a}
+
+/\V?\v{3,4}/8
+    >\x09\x20\x{a0}X\x0a\x0a\x0a<
+ 0: X\x{0a}\x{0a}\x{0a}
+
+/\H\h\V\v/8
+    X X\x0a
+ 0: X X\x{0a}
+    X\x09X\x0b
+ 0: X\x{09}X\x{0b}
+    ** Failers
+No match
+    \x{a0} X\x0a   
+No match
+    
+/\H*\h+\V?\v{3,4}/8 
+    \x09\x20\x{a0}X\x0a\x0b\x0c\x0d\x0a
+ 0: \x{09} \x{a0}X\x{0a}\x{0b}\x{0c}\x{0d}
+    \x09\x20\x{a0}\x0a\x0b\x0c\x0d\x0a
+ 0: \x{09} \x{a0}\x{0a}\x{0b}\x{0c}\x{0d}
+    \x09\x20\x{a0}\x0a\x0b\x0c
+ 0: \x{09} \x{a0}\x{0a}\x{0b}\x{0c}
+    ** Failers 
+No match
+    \x09\x20\x{a0}\x0a\x0b
+No match
+     
+/\H\h\V\v/8
+    \x{3001}\x{3000}\x{2030}\x{2028}
+ 0: \x{3001}\x{3000}\x{2030}\x{2028}
+    X\x{180e}X\x{85}
+ 0: X\x{180e}X\x{85}
+    ** Failers
+No match
+    \x{2009} X\x0a   
+No match
+    
+/\H*\h+\V?\v{3,4}/8 
+    \x{1680}\x{180e}\x{2007}X\x{2028}\x{2029}\x0c\x0d\x0a
+ 0: \x{1680}\x{180e}\x{2007}X\x{2028}\x{2029}\x{0c}\x{0d}
+    \x09\x{205f}\x{a0}\x0a\x{2029}\x0c\x{2028}\x0a
+ 0: \x{09}\x{205f}\x{a0}\x{0a}\x{2029}\x{0c}\x{2028}
+    \x09\x20\x{202f}\x0a\x0b\x0c
+ 0: \x{09} \x{202f}\x{0a}\x{0b}\x{0c}
+    ** Failers 
+No match
+    \x09\x{200a}\x{a0}\x{2028}\x0b
+No match
+     
+/a\Rb/I8<bsr_anycrlf>
+Capturing subpattern count = 0
+Options: bsr_anycrlf utf8
+First char = 'a'
+Need char = 'b'
+    a\rb
+ 0: a\x{0d}b
+    a\nb
+ 0: a\x{0a}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    ** Failers
+No match
+    a\x{85}b
+No match
+    a\x0bb     
+No match
+
+/a\Rb/I8<bsr_unicode>
+Capturing subpattern count = 0
+Options: bsr_unicode utf8
+First char = 'a'
+Need char = 'b'
+    a\rb
+ 0: a\x{0d}b
+    a\nb
+ 0: a\x{0a}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    a\x{85}b
+ 0: a\x{85}b
+    a\x0bb     
+ 0: a\x{0b}b
+    ** Failers 
+No match
+    a\x{85}b\<bsr_anycrlf>
+No match
+    a\x0bb\<bsr_anycrlf>
+No match
+    
+/a\R?b/I8<bsr_anycrlf>
+Capturing subpattern count = 0
+Options: bsr_anycrlf utf8
+First char = 'a'
+Need char = 'b'
+    a\rb
+ 0: a\x{0d}b
+    a\nb
+ 0: a\x{0a}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    ** Failers
+No match
+    a\x{85}b
+No match
+    a\x0bb     
+No match
+
+/a\R?b/I8<bsr_unicode>
+Capturing subpattern count = 0
+Options: bsr_unicode utf8
+First char = 'a'
+Need char = 'b'
+    a\rb
+ 0: a\x{0d}b
+    a\nb
+ 0: a\x{0a}b
+    a\r\nb
+ 0: a\x{0d}\x{0a}b
+    a\x{85}b
+ 0: a\x{85}b
+    a\x0bb     
+ 0: a\x{0b}b
+    ** Failers 
+No match
+    a\x{85}b\<bsr_anycrlf>
+No match
+    a\x0bb\<bsr_anycrlf>
+No match
+ 
+/ End of testinput 8 / 
diff --git a/lib/stdlib/test/run_pcre_tests.erl b/lib/stdlib/test/run_pcre_tests.erl
index 0ef3986918..8c6424e708 100644
--- a/lib/stdlib/test/run_pcre_tests.erl
+++ b/lib/stdlib/test/run_pcre_tests.erl
@@ -1,19 +1,19 @@
 %%
 %% %CopyrightBegin%
-%% 
-%% Copyright Ericsson AB 2008-2009. All Rights Reserved.
-%% 
+%%
+%% Copyright Ericsson AB 2008-2010. All Rights Reserved.
+%%
 %% The contents of this file are subject to the Erlang Public License,
 %% Version 1.1, (the "License"); you may not use this file except in
 %% compliance with the License. You should have received a copy of the
 %% Erlang Public License along with this software. If not, it can be
 %% retrieved online at http://www.erlang.org/.
-%% 
+%%
 %% Software distributed under the License is distributed on an "AS IS"
 %% basis, WITHOUT WARRANTY OF ANY KIND, either express or implied. See
 %% the License for the specific language governing rights and limitations
 %% under the License.
-%% 
+%%
 %% %CopyrightEnd%
 %%
 -module(run_pcre_tests).
@@ -25,7 +25,7 @@ test(RootDir) ->
     erts_debug:set_internal_state(available_internal_state,true),
     io:format("oldlimit: ~p~n",[ erts_debug:set_internal_state(re_loop_limit,10)]),
     Testfiles0 = ["testoutput1", "testoutput2", "testoutput3", "testoutput4",
-		 "testoutput5", "testoutput6", "testoutput10"],
+		 "testoutput5", "testoutput6","mod_testoutput8","testoutput10"],
     Testfiles = [ filename:join([RootDir,FN]) || FN <- Testfiles0 ], 
     Res = [ begin io:format("~s~n",[X]), t(X) end || X <- Testfiles ],
     io:format("limit was: ~p~n",[ erts_debug:set_internal_state(re_loop_limit,default)]),
@@ -42,12 +42,14 @@ t(OneFile,Num) ->
     put(error_limit,Num),
     put(skipped,0),
     Res = 
-	[test(Structured,true,index),
-	 test(Structured,false,index),
-	 test(Structured,true,binary),
-	 test(Structured,false,binary),
-	 test(Structured,true,list),
-	 test(Structured,false,list)],
+	[test(Structured,true,index,false),
+	 test(Structured,false,index,false),
+	 test(Structured,true,index,true),
+	 test(Structured,false,index,true),
+	 test(Structured,true,binary,false),
+	 test(Structured,false,binary,false),
+	 test(Structured,true,list,false),
+	 test(Structured,false,list,false)],
     {lists:sum(Res),length(Structured)*6,get(skipped)}.
 
 
@@ -63,11 +65,21 @@ pick_exec_options([Opt|T]) ->
 pick_exec_options([]) ->
     {[],[]}.
 
-test([],_,_) ->
+test([],_,_,_) ->
     0;
-test([{RE,Line,Options0,Tests}|T],PreCompile,XMode) ->
+test([{RE0,Line,Options0,Tests}|T],PreCompile,XMode,REAsList) ->
     %io:format("."),
     %case RE of <<>> -> io:format("Empty re:~w~n",[Line]); _ -> ok end,
+    Unicode = lists:member(unicode,Options0),
+    RE = case REAsList of
+	     true ->
+		 if 
+		     Unicode -> unicode:characters_to_list(RE0);
+		     true -> binary_to_list(RE0)
+		 end;
+	     false ->
+		 RE0
+	 end,
     {Options,ExecOptions} = pick_exec_options(Options0),
     {Cres, Xopt} = case PreCompile of
 		       true ->
@@ -80,7 +92,7 @@ test([{RE,Line,Options0,Tests}|T],PreCompile,XMode) ->
 	    %erlang:display({testrun,RE,P,Tests,ExecOptions,Xopt,XMode}),
 	    case (catch testrun(RE,P,Tests,ExecOptions,Xopt,XMode)) of
 		N when is_integer(N) ->
-		    N + test(T,PreCompile,XMode);
+		    N + test(T,PreCompile,XMode,REAsList);
 		limit ->
 		    io:format("Error limit reached.~n"),
 		    1;
@@ -91,12 +103,12 @@ test([{RE,Line,Options0,Tests}|T],PreCompile,XMode) ->
 			_ ->
 			    put(skipped,1)
 		    end,
-		    test(T,PreCompile,XMode)
+		    test(T,PreCompile,XMode,REAsList)
 	    end;
 	{error,Err} ->
 	    io:format("Compile error(~w): ~w~n",[Line,Err]), 
 	    case get(error_limit) of
-		infinite -> 1 + test(T,PreCompile,XMode);
+		infinite -> 1 + test(T,PreCompile,XMode,REAsList);
 		X ->
 		    case X-1 of
 			Y when Y =< 0 ->
@@ -104,7 +116,7 @@ test([{RE,Line,Options0,Tests}|T],PreCompile,XMode) ->
 			    1;
 			Y ->
 			    put(error_limit,Y),
-			    1 + test(T,PreCompile,XMode)
+			    1 + test(T,PreCompile,XMode,REAsList)
 		    end
 	    end
     end.
@@ -549,6 +561,8 @@ tr_option($N) ->
     [no_auto_capture];
 tr_option($8) ->
     [unicode];
+tr_option($U) ->
+    [ungreedy];
 tr_option($g) ->
     [{exec_option,g}];
 tr_option(_) ->