Tic‑Tac‑Toe — Flutter (Supabase auth + AI opponent)

This repository contains a Flutter Tic‑Tac‑Toe app that demonstrates:

• Email/password authentication using Supabase (sign-in and sign-up).
• Single‑player gameplay against a configurable AI (easy, medium, hard).
• Cross‑platform support (Android, iOS, web, desktop) using Flutter.

Project structure (important files)

• lib/main.dart — App entry point. Contains:
  • Supabase initialization.
  • Login screen (email/password).
  • GamePage: UI for the Tic‑Tac‑Toe board, game state, and controls.
  • AI integration via chooseAIMove (minimax algorithm).
• lib/game/ai.dart — (if present) alternate/extended AI implementation.
• pubspec.yaml — Dependencies (make sure supabase_flutter is present).
• android/, ios/, web/, windows/, macos/, linux/ — Platform-specific project files.

How the app works (high level)

1. main()

   • Ensures Flutter bindings are initialized.
   • Initializes Supabase client with project URL and anon key.
   • Starts the Flutter app.

2. Login UI

   • Validates email & password.
   • Calls Supabase auth:
     • Attempt sign-in via signInWithPassword.
     • If sign-in fails (no session/user), attempts signUp.
   • On success, navigates to GamePage with player email as identifier.

3. GamePage

   • Local state: _board (List), _current player ('X'|'O'), _winner, _moves, _winningLine.
   • Player taps a cell to place a mark. After a player move:
     • Check winner via _checkWinner.
     • If game not finished and it's AI's turn, call _performAIMove().

4. AI (chooseAIMove)

   • Easy: random empty cell.
   • Medium: runs limited-depth minimax plus a 30% chance of a random move to simulate mistakes.
   • Hard: full-depth minimax for perfect play.
   • Minimax details:
     • Board converted to ints: X=1, O=-1, empty=0.
     • Terminal states return scores: positive when AI wins (score = 10 - depth), negative when opponent wins (depth - 10), and 0 for draws.
     • Move ordering heuristic: center > corners > sides to guide search and improve pruning effectiveness (no alpha-beta implemented in this variant).
     • Returned index is validated before applying; AI actions are separated from UI thread via async delays to show “thinking”.

How the AI was coded (technical)

• Conversion helper: map board strings to integers to compute lines quickly.
• Winner check: iterate the 8 winning lines and compare three positions for equality and non-empty.
• Minimax function:
  • Recursively attempts each available move, flips current player, and increases depth.
  • Stops recursion when:
    • A terminal winner is found.
    • The board is full (draw).
    • Reached maxDepth (used for medium difficulty to limit search).
  • Scores incorporate depth to prefer faster wins and slower losses.
• Randomness for difficulty:
  • easy: purely random choice among empties.
  • medium: 30% probability to choose a random move to simulate non-perfect play.
  • hard: no randomness (perfect play at full depth).

Setup — local development (Android)

1. Ensure Flutter is installed and configured for Android.
2. Edit lib/main.dart (or create a lib/config.dart) and set:
   • SUPABASE_URL = your Supabase project URL.
   • SUPABASE_ANON_KEY = your Supabase anon (public) key. IMPORTANT: For production, do not commit keys. Use secure secrets or platform-specific environment config.
3. Add dependency in pubspec.yaml:

   dependencies:
     flutter:
       sdk: flutter
     supabase_flutter: ^1.4.0

4. From project root (where pubspec.yaml lives):
   • flutter pub get
   • flutter run # or flutter run -d emulator-5554
   • flutter build apk # to produce an Android APK

Supabase configuration

• In the Supabase dashboard:
  • Go to Authentication → Settings and enable Email/Password provider if required.
  • The anon key must have permissions for auth endpoints (standard).
• For debugging authentication errors, check:
  • Project URL and anon key are correct.
  • Network access from device/emulator.
  • Supabase project has Email provider enabled.

Troubleshooting tips

• "Couldn't resolve package 'supabase_flutter'":
  • Ensure pubspec.yaml is in the same folder as lib/.
  • Run flutter clean then flutter pub get.
  • Restart your editor/IDE if the analyzer fails to update.
• Null errors for Supabase:
  • Ensure Supabase.initialize(...) runs before runApp() and pub packages are resolved.
• Debugging auth:
  • Use ScaffoldMessenger snackbars (already present) to view Supabase exception messages.

Extending & production notes

• Replace in-source keys with environment variables or platform secure storage.
• Add persistent user profiles in Supabase (e.g., store statistics in a matches table).
• Consider adding alpha‑beta pruning to speed up the minimax search for larger boards or more complex heuristics.
• Add unit tests for game logic (winner detection, move validation) — tests exist in test/ as examples.

How to present this tomorrow (suggested slide order)

1. Brief overview & goals.
2. Architecture diagram (main → auth → game → AI).
3. Live demo: login → play vs AI (switch difficulties).
4. Code walkthrough:
   • show chooseAIMove and explain minimax lines and scoring.
   • show _submit authentication flow.
   • show _checkWinner and UI highlights for winning line.
5. Deployment steps and security caveats.
6. Q&A and next steps.

---

If you want, I can:

• Create a dedicated lib/config.dart and move SUPABASE constants there and update .gitignore suggestions.
• Add short inline comments in main.dart for the specific functions you will highlight during the presentation.
• Generate a one‑page slide text outline you can copy into PowerPoint/Google Slides.