AUTOMATED FRONT-END TO BACK-END REQUEST CAPTURING SYSTEM FOR FLUTTER-BASED APPLICATIONS

Abstract

Understanding and documenting connections within a software project can be challenging for both developers and non-technical stakeholders. To address this, we developed a Python script for automated request capturing and automated documentation of scripts in a Flutter-based project using its source code. The script pulls code from Bitbucket based on the client's name and captures key details such as class names, methods, HTTP methods, parameters, and request bodies using Regex. It supports various connections like front-end App to Middleware (both client-side and server-side), Middleware to ERP, and back-end App to Middleware services in languages such as Dart and Java. The results are compiled into a YAML file and displayed on Swagger UI for easy access. Initially, the script included AI-generated descriptions using ChatGPT's GPT-3.5 Turbo model, but this feature was removed due to rate limitations. The entire setup is containerized with Dockerization and integrated into Commsult's customer portal, the Customer Portal, where it can be triggered by a button click that activates an Express.js server. This tool has proven effective in helping both developers and non-technical users understand and navigate the project's codebase, despite the removal of the LLM-generated descriptions.