The description of Tk_PointToChar in TextLayout(3)
seems to indicate that the return value is always a
valid character index (except in the special case of an
empty string). It can actually return an index *one
past* the end of the string (even when the string is
nonempty). Manpage should be clarified to reflect this.
Fixed in HEAD. Can't be bothered to backport such a small docfix.