For a long time it has been a common limitation that there are no open source applications that are able of reading data in SAS data sets, aka. sas7bdat files. This format, commonly used in the SAS statistical software system, is a pretty intricate binary format which until now was a black box for open source tools to handle. Effectively this meant that if you had SAS data you also needed SAS software - even just to read it.

SassyReader is a Java-based reader of the binary format. It is not a connector to any SAS systems or the like, rather it reads the raw data and exposes it as Java objects. This makes life a lot easier and adoption is expected soon within the open source community of data-oriented applications (BI solutions, ETL tools and the like).

The website contains a few code examples: http://sassyreader.eobjects.org/

SassyReader is already available in the central Maven repositories, so including it in your projects should be quite easy.

The founder of the project, Kasper Sørensen, who have also founded project like DataCleaner and MetaModel, also has a write-up on his blog about the motivations and enthusiasm of creating the library: http://kasper.eobjects.org/

As the project is a reverse engineered library based on the raw binary format of the files there are still some limitations to it's capabilities. SAS data sets vary according to the operating systems they where written with and other factors, but in general it works on most SAS data sets. There is still room for improvement and interested developers are invited to join the development.