Introduction

The purpose of this document is to give a clear explanation of the scanner and parser for abc. In defining the scanner and lexer our goal was to come up with clear rules for tokens and to express the grammar in an LALR(1) specification that results in no shift-reduce or reduce-reduce conflicts. Any language extensions implemented using abc should follow the same principles.

We have based our implementation on Polyglot [1], starting with the base Java compiler. The scanner is an extended version of polyglot's base Java scanner. The main difference is that the abc scanner uses states to distinguish between different contexts. The parser uses polyglot's mechanism for extending an existing grammar.

The actual code specifying the scanner and parser can be found in the directory abc/src/abc/aspectj/parse in the files aspectj.flex and aspectj.ppg.