| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 581.39 KB | Adobe PDF |
Advisor(s)
Abstract(s)
This thesis explores the implementation of clean code best practices in Golang, a language
designed for performance and efficiency. Emvenci is transitioning its codebase to a new architecture,
adhering to modern software design principles. The primary objective is to refactor and
optimize all e-learning service endpoints under this new architecture, with a focus on enhancing
performance through parallelization and concurrency. A secondary objective is to propose additional
guidelines, as needed. During the implementation, a systematic routine was established
to streamline the process, documenting key steps and highlighting improvements to aid testers
who lack direct access to the code. Performance optimization techniques were carefully applied to
maintain a balance between complexity and efficiency. These techniques included parallelizing independent
database accesses, consolidating repeated queries, and eliminating redundant code and
unnecessary processing. Testing was conducted in two phases: clean code analysis and performance
evaluation. Clean code metrics, such as maintainability index, cyclomatic complexity, and
cognitive complexity, were assessed. In the performance phase, load tests measured processing
time, CPU, and memory usage across endpoints. The results demonstrated significant improvements
in system performance, though clean code metrics showed a decline due to the increased
volume and complexity of the refactored functions. While the new architecture initially risked
degrading performance, the applied optimizations ensured improvements in most endpoints. This
suggests that similar techniques may yield benefits when refactoring other parts of the code. Additionally,
the newly established documentation for both the clean architecture and performance
testing will provide a foundation for future work at the company, setting expectations for quality
and consistency.
Description
Trabalho de projeto de mestrado, Informática, 2025, Universidade de Lisboa, Faculdade de Ciências
Keywords
Código limpo Otimização GoLang Melhores práticas Desenvolvimento de Software Teses de mestrado - 2025