AutoMoose offers an integration of Moose with the command line.
The diagram below offers an schematic overview of the process followed by Auto Moose. The first step is to get inFusion parse the external sources (typically Java). Alternatively, if other kinds of sources are targeted, it is possible to use directly Moose or another parser. The parser creates the model and stores it in an MSE file.
The Moose image is started and it is passed several scripts to execute. These scripts are typically one liners Smalltalk commands that trigger various computations inside the Moose image. For example, a default scripts is simply loading the MSE file in Moose, while another script is spawning the MoosePanel.
Some scripts are already provided by default, but others can be built to match the contextual situation in which the report is desired. For example, a possible script could be to trigger dedicated analysis, or spawn a dedicated report.
Finally, after the image is prepared, it is copied together with the MSE file and with the sources into an archive.
AutoMoose comes with a built in compatibility with the Hudson continuous integration server, and can simply be integrated with others. See Section 23.2 for more information.