... as before, I'm using utf8 or euc encoded strings...

Okay, how do you know when you're using one or the other? Does the web site (play-asia.com) support both encodings, and if so, how do you tell it which one you're using?

ord returns the numeric representation of each byte, not each character. The above gives 6 distinct codes, not 3.

6 numerics instead of 3 for that 3-character string definitely means "not utf8" (so presumably euc, based on what you've said); if the web site is looking for utf8-encoded numerics, you should use Encode to convert from euc to utf8, then use ord():

use Encode;
...

my $string = "...";  # wherever you get your euc string from;
$string = decode( "euc-jp", $string );  # now it's a utf8 string;

... # plug it into the url as described earlier
[download]

What does "rovf" mean, and am I missing your point(s) completely? (sorry-- I just realized that "rovf" was another monk -- moving right along...)

Your descriptions and replies are a bit hard for me to follow. And what do you mean by "the values are actually just HTML representation"? There are numeric character entity references (both decimal and hex, used in HTML, XML, SGML), there are symbolic entity references (like "á"), and there are uri-encoded versions of these (with the punctuation marks converted to "%" followed by two hex digits).

Are you really still having a problem with this?

Only if perl doesn't know that the string really is utf8. If the string is coming from a data file, and you open that file without telling perl specifically what type of encoding to use for it, perl reads data from the file as bytes, and you need to decode() the string (using the Encode module) so that perl will see it properly converted/interpreted as utf8 characters, and the scalar variable holding the string will be flagged as such. Note the following:

# open an euc-jp file:
open( $inp, "<:encoding(euc-jp)", "some_file.euc" ) or die "$!";

while (<$inp>) {  # data will be decoded on input
    ... # so $_ will have its utf8 flag "on" and 
        #   s///, ord(), etc will use character semantics
}

# open a utf8 file:
open( $inp, "<:utf8", "some_file.utf8" ) or die "$!";

while (<$inp>) {  # data will be interpreted as utf8 on input
    ... # so $_ will have its utf8 flag "on" and... (same as above)
}
[download]

If your string is based on some literal value in your perl script file, you have to make sure that the editor you use to compose the source code is saving the script as a utf8 file (and you probably need to put use utf8; in the script, so perl knows how to interpret the literal utf8 characters that it contains).

C:\>perl -e"print unpack q!H*!, chr 333"
c58d
C:\>perl -e"print ord chr 333"
333
C:\>perl -e"print chr 333"
Wide character in print at -e line 1.
+ì
C:\>perl -MEncode -e"print ord decode q!UTF-8!,encode q!UTF8!,chr 333"
333
[download]