Since I am reading the names from Excel, so I assumed the source(text of name) is sure UTF-8

Well, the docs don't agree with your assumption, see http://search.cpan.org/perldoc/Win32::OLE#CP, you need to use CP_UTF8

I had a little study. And I tried Win32API::File::MoveFile() and Win32::Unicode::File::moveW() but without success. I had also tried to create a batch file and run it in cmd, still not success.

And the error message was? Maybe you need to post your code :)

Thanks for reply. Um.. to narrow down the err possibility, I've cut and paste the 2 cols into a text file, use notepad, and save as UTF-8. But please tell me if that still not the version that Perl happy with. Here's my code and there are 2 paths see the remarked lines as version 2 :

use Win32API::File 0.08 qw( :ALL );
#use Win32::Unicode::File;
binmode STDOUT, ":utf8";

open F, "Result_fineName.txt";
#binmode F, ":utf8"; # binmode or not still won't work
my @list = <F>;
close F; chomp @list;

my %hash;
foreach (@list) {
    my ( $id, $name ) = split /\t/, $_, 2;
    $name =~ s/[\\\/\<\>\?\:\|\*]/ /g;
    $name =~ s/ {2,}/ /g;
    $hash{$id} = $name;
}


chdir ( "./doRename/2.4" ) ;
my $source = `dir /b`; 
my @src = split /\n/, $source;
chomp @src;

foreach my $f ( @src ) {
    my @parts = split /[ _]/, $f;
    print $f ;
    MoveFile $f, "$hash{$parts[1]}.mp3" or print  fileLastError() ; #W
+in32API::File
        #moveW $f, "$hash{$parts[1]}.mp3" or print errorW; # Win32::Un
+icode::File

    print $/;    
}
[download]

There are 2 sample files in folder 2.4.
One will rename to a file name starts with ANSI char. If I use the Win32API::File way, there is no error raised, it do the rename, but it turns into monster char.
If I use the Win32::Unicode::File way, it raised an err Undefined subroutine Errno::ERROR_FILE_EXISTS called at C:/Perl/site/lib/Unicode/Error.pm line 31.


The other will rename to a pure Japanese file name.
If I use the Win32API::File way, it rasied an err : The filename, directory name, or volume label syntax is incorrect
If use the Win32::Unicode::File way, it does nothing, no err, and no rename.

Note, before I run this from my cmd prompt, I've changed codepage to 65001. Without this, the error message will raise in Chinese ( I guess ), as it showed in monster as well.

.. to narrow down the err possibility, I've cut and paste the 2 cols into a text file, use notepad, and save as UTF-8.... as it showed in monster as well.

???

moveW works for me, start cmd , chcp 65001, perl kebab.pl > keb.txt , type keb.txt, notepad keb.txt

#!/usr/bin/perl --
use strict; use warnings;
use Path::Class;
use File::Slurp;
use Data::Dump;
use Win32::Unicode::File;

our  $thisf = file(__FILE__)->absolute;
our  $thisd = $thisf->dir;
our  $tmp   = 'fafafa';
chdir $thisd or die sprintf q/ERRRR(%d)(%s)(%d)(%s)/, $!,$!,$^E,$^E;

my $names = file('utfnamelist.txt')->absolute;
#~ my $bin = read_file( $names, { binmode => ':raw' } ) ; dd $bin;
my $bin = "\xEF\xBB\xBF# a utf-8 file-o kebab's\r\nutfnamelist.txt    
+\xD1\x9B\xD0\xB5\xD0\xB2\xD0\xB0\xD0\xBF.txt\r\n\xD1\x9B\xD0\xB5\xD0\
+xB2\xD0\xB0\xD0\xBF.txt          ra\xC5\xBEnji\xC4\x87.txt\r\nra\xC5\
+xBEnji\xC4\x87.txt        \xC4\x87evap.txt\r\n\xC4\x87evap.txt       
+   \xD0\xBA\xD0\xB5\xD0\xB1\xD0\xB0\xD0\xBF\xD1\x87\xD0\xB5.txt\r\n\x
+D0\xBA\xD0\xB5\xD0\xB1\xD0\xB0\xD0\xBF\xD1\x87\xD0\xB5.txt        keb
+ab.txt\r\n";

mkdir $tmp  or die sprintf q/ERRRR(%d)(%s)(%d)(%s)/, $!,$!,$^E,$^E;
chdir $tmp  or die sprintf q/ERRRR(%d)(%s)(%d)(%s)/, $!,$!,$^E,$^E;

write_file( $names->basename, \$bin );

binmode STDOUT, ':encoding(UTF-8)';print "\x{feff}";#BOM

ddir();

open my($fh), '<:encoding(UTF-8)', $names or die sprintf q/ERRRR(%d)(%
+s)(%d)(%s)/, $!,$!,$^E,$^E;
while(<$fh>){
    next if $. == 1; # /#/;
    my( $here, $there ) = split /\s+/, $_;
    next unless $here and $there;
    print "# moveW  $here   =>    $there\n";
#~     dd [ $here, $there ];
    moveW( $here, $there )  or die sprintf q/ERRRR(%d)(%s)(%d)(%s)/, $
+!,$!,$^E,$^E;
    ddir();
}

chdir '..';
dir($tmp)->rmtree("verbose","cautious");

sub ddir {
    my $dir = `cmd /x /c "dir /b "`;;
#~     use Encode::Detective qw' detect                          ';
#~     warn detect($dir) ; # "UTF-8"
    utf8::decode($dir); 
    print "# dir # ", $dir ;
}
__END__
[download]

# dir # utfnamelist.txt
# moveW  utfnamelist.txt   =>    ћевап.txt
# dir # ћевап.txt
# moveW  ћевап.txt   =>    ražnjić.txt
# dir # ražnjić.txt
# moveW  ražnjić.txt   =>    ćevap.txt
# dir # ćevap.txt
# moveW  ćevap.txt   =>    кебапче.txt
# dir # кебапче.txt
# moveW  кебапче.txt   =>    kebab.txt
# dir # kebab.txt
unlink fafafa\kebab.txt
rmdir fafafa