Social Network Simulation

Author: Harsh Raj Singh (250847420)

Overview

This project implements a concurrent social network simulation using Advanced Functional Programming techniques. It models 10 independent user threads interacting via message passing, designed to demonstrate High-Assurance software architecture.

Unlike standard implementations, this project uses:

• Software Transactional Memory (STM) for deadlock-free concurrency (replacing raw MVars).
• Tagless Final style to decouple pure logic from IO effects.
• Type-Level States (GADTs) to make invalid agent states unrepresentable.

Extensions Implemented (Top 20%)

1. Mood Propagation Dynamics: Users are emotional agents. Their internal mood (Happy, Neutral, Sad) evolves based on the "Vibe" of messages received, creating a dynamic feedback loop.
   • Effect: Mood influences both the content of sent messages and the response latency (simulating human reaction times).
2. Log Parsing & Analytics: A custom Megaparsec grammar parses the simulation logs to generate an ASCII Pie Chart of the final network mood distribution, verifying the propagation mechanics.

Usage

The application uses a CLI for various operations:

1. Run Simulation (Default):

   stack run

   Executes the simulation (10 users, 100 messages) and prints the final message counts.

2. Analyze Log (Visualization):

   stack run -- analyze-log

   Parses the generated social_network_chat.log and displays the Mood Distribution Pie Chart.

3. View Log/Help:

   stack run -- view-log
   stack run -- help

Testing & Verification

The project employs Property-Based Testing (QuickCheck) to verify invariants (e.g., algebraic laws for counters, round-trip correctnes for parsers).

• Run All Tests:

  stack test

• Check Code Coverage:

  stack test --coverage

  (Open the generated HTML report in .stack-work/.../hpc/index.html to verify 100% coverage of pure logic modules).

Project Structure

• app/Main.hs: CLI Entry point.
• src/Run.hs: Simulation wiring and Supervisor.
• src/Config.hs: Global simulation constants.
• src/UserBehaviour.hs: Core "Brain" logic (Pure).
• src/Types.hs: Domain types (User, Message, Monoids).
• src/Env.hs: Runtime environment context.
• src/Effects.hs: Abstract Effect capability definitions.
• src/AppM.hs: Concrete interpreter (ReaderT Env IO).
• src/Logger.hs: Concurrent logging service.
• src/LogParser.hs: Parsing logic for analytics.
• test/Spec.hs: Test suite (Properties & Unit tests).