Size/lookahead tradeoff for LL(k)-grammars

The Phenomenon of Non-Recursive Trade-Offs

“…For example, LL(k +1) grammars are known to describe strictly more languages than LL(k) grammars, i.e., the length of the lookahead induces an infinite hierarchy. Nevertheless, the trade-offs between the levels of the hierarchy are recursive [2]. On the other hand, we have seen that there are non-recursive trade-offs between the hierarchy levels of ambiguous and nondeterministic pushdown automata.…”

“…Malcher [22] has investigated the trade-offs between (k + l)-turn and A;-turn pushdown automata. The results are summarized as follows: (1) between 1-turn pushdown automata and finite automata, (2) between (k + l)-turn pushdown automata and k-turn pushdown automata, (3) between finite-turn pushdown automata and k-turn pushdown automata, (4) between pushdown automata and finite-turn pushdown automata.…”

“…Now we extend the language in order to meet our purposes. Basically, the idea is to keep the structure of the language but to build the subwords Wj over an alphabet of pairs, that is, Wi = "* " 2 Proof. Let n = ( 2 ) +1-We sketch the construction.…”

Remembering Chandra Kintala

“…A "limited approach" to this trade-off was taken in [13] by introducing and investigating various delay-spectra where the parsers become gradually smaller as the maximal delay in recognizing an error slowly increases. And finally looking at parsers for LL(k) and LR(k) languages, the effect on the parser-size by gradually increasing or decreasing the look-ahead was studied in [1,2,57].…”