Syntax-driven data processing for practice

The article is devoted to the different definitions of syntax and semantics of programming languages on the basis of attribute approach. The syntactic description of attributes in VW and Koster affix grammars are compared for tasks of translations. The method of syntax definition of CS language is presented by means of the syntactic graph-scheme (SGS) constructed on regular expressions. To optimize a parser into simpler class of grammar the principle of grammar regularization has been shown, it uses the SGS equivalent transformations. The algorithm of one of equivalent transformations for the CS-grammar in a regular form has been given.