Methodology for automated information processing using a high-level programming language. Part 2

This part of the paper is devoted to the development of a language analyzer defined by translation grammars in a regular form. The described technique was first applied in the 1970s in a compiler for ALGOL 68. Constraints are formulated on the grammar, presented in the form of a system of multicomponent directed graphs (like extended Wirth diagrams), guaranteeing the existence of a language processor similar to a deterministic pushdown transducer, which is then considered as a control mechanism scanning the input chain from left to right, fixing the sequence of states to initiate actions that make up the translation process. The specificity of the technique is the grammar regularization algorithms based on equivalent transformations of the grammar syntactic graph: elimination of recursion and insertion of iterations. Regularization of the original grammar is a part of the full cycle of language implementation, consisting of the user cycle and the semi-automated developer cycle.