Decomposes Tibetan characters in the buffer into their components. See also the documentation of the function `tibetan-decompose-region'. (fn)