The Netherlands.

From the point of view of parsing the LL(K] grammars constitute a very attractive class of context-free grammars. For each LL{K] grammar a top-down parsing algorithm can be devised which is essentially a one-state deterministic push-down automaton. From a more theoretiG point of view LL[K) grammars are attractive as well. It is well-known, for example, that it is decidable whether two LLfk) grammars are equivalent. Also the hierarchy of LL[K) languages with regard to the length k of the look-ahead is a characteristic property. The class of LL[k] grammars is properly contained in the class of LR[K] grammars, and even the family of LLCK) languages is properly contained in the family of LR(K) languages. If we focus on the "gap " between LL~K) and LR[K] grammars the following points are of interest, [i] There is the obvious difference in grammar definition. (ii) The generating capacities are different. (iii) Apart From the difference between LR[O) and LR(1) languages the length k of the look-ahead does not play a role for LR[K] languages