Gawk and multi-byte characters

Hi All,

I am working with UTF-8 encoded files and wonder how Gawk is treating 
multibyte characters. Here an example:

$ cat AAA.txt
AAA
ÅÃÆ

The length() function counts 3 characters in both records:

$ awk '{print length()}' AAA.txt
3
3

But printf seems to count bytes, rather than characters

$ awk '{printf "%-7s|\n", $0}' AAA.txt
AAA    |
ÅÃÆ |


What about other functions: index(), substr(), etc.? Is this do***ented 
somewhere?

TIA. Hermann