From Wikipedia: :-)
Cyrillic Extended A (2DE0...2DFF) and Cyrillic Extended B (A640...A69F)
"Arabic" is a bit more complicated. It seems some characters are shared: 06D5, for example, is listed on Unicode.org as "ARABIC LETTER AE • Uighur, Kazakh, Kirghiz", and other characters in these languages appear scattered through the Arabic basic set (0600..06FF). The Extended Arabic set (0750..077F) contains "miscellaneous" characters -- as few as only two additional ones for Torwali.
(See
http://unicode.org/charts/ -- although it deals with
scripts rather than languages. FWIW, I agree with your use of "language" rather than "script", inasmuch it sounds a bit more obvious to the casual inter-linguistic typograph.)
Perhaps you could start with Unicode basic blocks, then assign single characters to the defined languages and throw everything else in that block into an "Other" tab, similar to the Library section?
