Examee: An Internal DSL-based framework for composing personalized exams in a .NET environment.

Abstract

This thesis presents Examee, a dynamic platform designed to allow teachers to create, customize, and build exams through a custom domain-specific internal language (DSL) embedded in a .NET library. The Examee tool set leverages an internal DSL, allowing teachers to define exam structures, add custom parameters, and configure exam content exactly as required. Teachers create exam templates as .DLL (Dynamic Link Library) files using Examee's .NET-based library, incorporating a rich set of design templates, including Builder, Composite and Intermediate Representation (IR) Tree, to facilitate complex, hierarchical definition of exam content. Once created, these templates can be uploaded to the Examee platform for automated production of unique exams. On the platform, the backend architecture, based on .NET Core, is integrated with a PostgreSQL database to manage user information, exam templates and query data extracted directly from uploaded DLLs. Utilizing additional design patterns such as Repository, Unit of Work and Dependency Injection, the backend securely executes these DLLs and extracts data for further processing. Exams are created dynamically based on the provided parameters, with the results saved in the desired PDF format and securely stored for user download. Examee's frontend, developed using Angular, provides an improved interface for uploading files, recording and managing exam downloads, using cloud storage solutions to manage file transfer.