Architecture

The Open Voice Factory’s architecture is effectively that of a compiler – with a python ‘front end’ that accepts a PowerPoint file and created an intermediate form, and a Javascript back-end that read the intermediate form and presents it to the user as a speech aid.

The python front end relies heavily on the open source library python-pptx, which we paid some code bounties to improve. The intermediate form that is created is (now) the Open Board Format, an (JSON-based) open source effort to improve speech aid compatibility that we are supporting. Our javascript is currently undergoing a large rewrite but is currently very low tech. Speech synthesis is provide by Google for the online version.

A description of the functionality of the system is here

The system is broken into two main components, one to convert pptx to obf, and the other to display obf and deliver the board functionality.

Upload Architecture:

• Physical/Deployment
• Logical
• Process

Display Architecture

• Physical/Deployment
• Logical
• Process

Note that the architecture adopts the principles of 4+1, although currently no Development view is provided as this is contained in the Physical and Logical.

The source architecture diagrams are created in draw.io, here