Design and implementation of a parser for Java 1.5 programming language

A parser is a program that parses a text written according to some specific rules and represents the text in memory with an appropriate data structure. Parsers can be used in a variety of areas in computer science. They can be used as an intermediate step in compilers that convert a programming language program into another programming language program. They are also extensively used in natural language processing.In this thesis, we try to design and implement a parser targeting at programs based on Java 1.5 specifications. We use JFlex and CUP software tools as an automatic scanner generator and an automatic parser generator. JFlex employs regular expressions to identify tokens. CUP uses context free grammars to specify the language's syntactic rules. It produces top-down LALR parsers via which most of the programming languages can be parsed. Our parser can parse multiple Java files and represent them as a single parse tree. To test if our parser runs correctly we apply the following steps. After parsing original Java files we traverse the parse tree and dump it into new Java files. These dumped files constitute our first group of files. Then the dumped Java files in the first group are fed into our parser and the corresponding parse tree is dumped into another set of files, which constitute our second group of files. We compare each Java file in the first group with the corresponding Java file in the second group. After extensive experiments, we see that the compared Java files have a verbatim text.