DR parsers: a generalization of LR parsers

A method is presented to construct a parser for a visual language whose specification can be done by a context-free grammar. The main idea is to allow a traditional LR parser to choose the next symbol to parse from a two-dimensional space. The positional grammar is defined, and some examples are given. The DR parser, along with the parsing algorithm and examples, is presented (`D´ stands for driven scanning of the input, and `R´ for rightmost derivation in reverse.) The construction of DR(0) items and simple DR parsing tables are shown. Considerations of ambiguity are given, along with the construction of a simple DR parser for the arithmetic expression grammar. The results of a practical implementation using the Yacc tool are given