I read perluniintro and tried the examples of "How Do I Convert Binary Data Into a Particular Encoding, Or Vi ce Versa?" with multibyte character. And I think the author confusing for "native string" or "bytes" ... I think. It will make sense with letters less than 128 code point, but when I tried with a letter like 'HIRAGANA LETTER A', it doesn't make sense. In short, examples seems to me, it is forgetting "encoding to bytes".

"A" is 0x41 for bytes and 0x41 for code point.
"HIRAGANA LETTER A is 0xe3,0x81,0x82 for bytes and 0x3042 for codepoint.

#hex dump of A
#00000000  41                                                |A|
#00000001
[download]

#hex dump of HIRAGANA LETTER A
#00000000  e3 81 82                                          |...|
#00000003
[download]

And two example codes below.

#Example 1: native string may not be native string
#Code: $native_string=pack('W*', unpack('U*', $unicode_string));

use strict;
use warnings;
use Encode qw(encode);
use Devel::Peek;
use 5.012;

    my($code_point,$unicode_string,$native_string, $native_string2);
    $code_point=0x41;#"A";
    $unicode_string=pack('U*', $code_point);
    $native_string=pack('W*', unpack('U*', $unicode_string));

    Dump $unicode_string;
    Dump $native_string; # ==> here it is not UTF-8 flagged

    $code_point=0x3042;#HIRAGANA LETTER A
    $unicode_string=pack('U*', $code_point);
    $native_string=pack('W*', unpack('U*', $unicode_string));
    $native_string2=Encode::encode('utf8', $unicode_string);

    Dump $unicode_string;
    Dump $native_string;  # ==> this is UTF8 flaged  may be transparen
+tly upgraded because code point > 255
    Dump $native_string2;
[download]

Devel::Peek shows $native_string is UTF8 flagged and $native_string2 is not UTF-8 flagged in case of HIRAGANA LETTER A.

#Example 2: it is not bytes, it is array of code point.
#Code: @bytes=unpack("C*", $unicode_string);
use strict;
use warnings;
use Encode qw(encode);
use 5.012;

    my($code_point,$unicode_string,@bytes);

    $code_point=0x41;#A
    $unicode_string=pack('U*', $code_point);
    @bytes=unpack("C*", $unicode_string);
    print join('|', @bytes), "\n";

    $code_point=0x3042;#HIRAGANA LETTER A
    $unicode_string=pack('U*', $code_point);
    @bytes=unpack("C*", $unicode_string);
    print join('|', @bytes), "\n";  #==>these are not bytes ,but array
+ of codepoints


    $code_point=0x3042;#HIRAGANA LETTER A
    $unicode_string=pack('U*', $code_point);
    @bytes=map{ sprintf("%X",$_) } unpack("C*", Encode::encode('utf8',
+$unicode_string));
    print join('|', @bytes), "\n";
[download]

So, I want to hear from monks suggestions, comments or "read this document", anything. I am now reading perlunicode.

regards.