Arabic encoding using Unix commands

Shell Programming and Scripting
1

I am using sed on Arabic file (utf-8 encoding) like bellow:

sed 's/./& /g' file

and all I get is:

1   ? ?   ? ? ? ? ? ? ? ? ? ?   ? ? ? ?   
? ? ? ? ? ?   ? ? ? ? ? ? ? ? ? ?

I tried change the LANG variable to

LANG=en_US.UTF-8

but I still get the same "?" output. What is the easiest way to change this without transliterating to English letters?

---------- Post updated at 10:54 AM ---------- Previous update was at 09:54 AM ----------

I solved it.
After setting LANG=en_US.UTF-8, just edit .profile and add these lines:

export LC_NUMERIC=C
export LC_ALL=C
export LESSCHARSET=utf-8
export LC_ALL=$LANG
2

For this to work, you must have been piping the output of sed to less or to a file that you were then paging using less.