Abstract
Source code metrics help to evaluate the quality of the code, for example, to detect the most complex parts of the program. When writing a system which calculates metrics, especially when it has to support multiple source code languages, the biggest problem which arises is the creation of parsers for each supported language. In this paper we suggest an unusual Open Source solution, that avoids creating such parsers from scratch. We suggest and explain how to use parsers contained in the Eclipse IDE as parsers that support contemporary language features, are actively maintained, can recover from errors, and provide not just the abstract syntax tree, but the whole type information of the source program. The findings described in this paper provide to practitioners a way to use Open Source parsers without the need to deal with parser generators, or to write a parser from scratch.