Setup & Usage Guide

This guide explains how to set up the Legislative Analysis system locally and how to use its features for analyzing Sri Lankan laws.

Prerequisites

Ensure you have the following installed on your system:

• Docker Desktop: For running the containerized application.
• Make: For executing build automation commands.
• Python 3.10+: (Optional) For running scripts directly if needed.

Quick Start

The entire system is containerized and orchestrated via docker-compose, abstracted behind simple Makefile commands.

1. Start the Cluster

To spin up both the FastAPI backend and the Next.js frontend, run:

make cluster-up

This command will:

1. Build the Docker images for the frontend and backend.
2. Start the containers in detached mode.
3. Ensure the SQLite database is initialized.

2. Access the Application

Once the cluster is up, open your web browser and navigate to:

http://localhost:3000

You will be greeted by the Legislative Analysis dashboard.

Using the System

Browsing Acts

The home page lists all analyzed acts. You can:

• Search: Filter acts by title or keyword.
• Filter: Select specific domains (e.g., Agriculture, Finance).
• View Details: Click on any act card to view its full analysis.

System Architecture in Action

When you browse the UI:

1. Frontend: The Next.js app fetches data from the backend APIs.
2. Backend: The FastAPI service queries the research.db SQLite database.
3. Data: The data you see was pre-processed using Google Gemini 2.0 Flash.

Shutting Down

To stop the services and ensure all data is safely dumped to disk:

make cluster-down

Note: This command automatically runs a data dump before stopping the containers, ensuring no analysis data is lost.

Troubleshooting

"Cluster not starting"

• Check if Docker Desktop is running.
• Ensure port 3000 (Frontend) and 8000 (Backend) are free.

"Database cache missing"

• If the UI is empty, you may need to re-run the extraction or ensure the data/research.db file exists.