# Prison Management System – Code Restructuring Project This project demonstrates **software re-engineering / code restructuring** on a Prison Management System written in Python. It contains two versions: ``` prison-management/ ├── legacy/ <- BEFORE: messy, single-file, code smells │ └── prison_system.py │ └── refactored/ <- AFTER: layered architecture, design patterns └── prison_app/ ├── __init__.py ├── cli.py (presentation layer) ├── models/ │ ├── __init__.py │ ├── inmate.py │ ├── staff.py │ └── visitor.py ├── services/ (business logic layer) │ ├── __init__.py │ ├── inmate_service.py │ ├── staff_service.py │ └── visitor_service.py ├── repositories/ (data access layer / DAO) │ ├── __init__.py │ ├── database.py │ ├── inmate_repository.py │ ├── staff_repository.py │ └── visitor_repository.py └── utils/ ├── __init__.py ├── constants.py └── exceptions.py ``` --- ## 1. Legacy version (`legacy/prison_system.py`) A single God class `PrisonSystem` that mixes together: - Database setup and raw SQL - Input validation with magic numbers (18, 50, 100, 10000…) - Business logic (risk level calculation) - Display/print logic - A CLI menu, all in `main()` **Code smells present (for your report):** | Smell | Where | |---|---| | God Class | `PrisonSystem` does DB + validation + logic + display | | Magic Numbers | `18`, `50`, `100`, `10000`, `10`, `5` hardcoded | | Duplicated Code | Inmate lookup repeated in `add_visitor()` and `get_inmate_info()` | | Long Method | `main()`, `get_inmate_info()` | | Mixed Responsibilities | Business rule (risk level) embedded in print statements | | No Layering | UI, logic, and DB all directly coupled | --- ## 2. Refactored version (`refactored/prison_app/`) Restructured into a **layered architecture**: - **models/** – plain data classes (`Inmate`, `Staff`, `Visitor`) using `@dataclass`. `Inmate.risk_level` and `Inmate.release_year()` extract business rules out of print statements. - **repositories/** – DAO pattern. All SQL isolated here. `Database` is a **Singleton**, so only one connection exists. - **services/** – business logic + validation (`InmateService`, `StaffService`, `VisitorService`). `VisitorService` reuses `InmateService.get_inmate()` instead of duplicating the lookup query. - **utils/constants.py** – all magic numbers replaced with named constants. - **utils/exceptions.py** – `ValidationError`, `NotFoundError` replace scattered `print()` + `return`. - **cli.py** – presentation layer only; delegates everything to services. **Refactoring techniques applied (map this to your report):** | Technique | Applied to | |---|---| | Extract Class | God class → Inmate/Staff/Visitor models, services, repositories | | Extract Method | Validation logic, risk level calculation | | Replace Magic Number with Constant | `constants.py` | | Introduce DAO Pattern | `repositories/` | | Singleton Pattern | `Database` | | Replace Error Code with Exception | `ValidationError`, `NotFoundError` | | Separation of Concerns / Layered Architecture | models / services / repositories / cli | --- ## 3. How to run in VS Code ### Prerequisites - Install **Python 3.10+** - Install the **Python extension** in VS Code (by Microsoft) ### Step-by-step 1. **Open the folder** - Open VS Code → `File > Open Folder` → select `prison-management/` 2. **Select Python interpreter** - `Ctrl+Shift+P` → "Python: Select Interpreter" → choose your Python 3 install 3. **Run the legacy version** - Open `legacy/prison_system.py` - Open a terminal: `Terminal > New Terminal` - Run: ```bash cd legacy python prison_system.py ``` 4. **Run the refactored version** - Open a terminal at the project root (`prison-management/`) - Run: ```bash cd refactored python -m prison_app.cli ``` - Important: run it as a **module** (`-m prison_app.cli`) from inside `refactored/`, so Python resolves the `prison_app` package imports correctly. Alternatively, on Windows PowerShell / Linux / macOS you can set `PYTHONPATH`: ```bash cd refactored PYTHONPATH=. python prison_app/cli.py ``` 5. **Database files** - Both versions create a local `prison.db` SQLite file in the directory you run them from. - Delete `prison.db` if you want to reset all data. --- ## 4. Suggested report structure 1. Introduction & objectives 2. Description of legacy system + identified code smells (table above) 3. Restructuring plan & techniques chosen 4. New architecture diagram (models → services → repositories → database) 5. Before/after code comparison (pick 2–3 methods, e.g. `add_inmate`, `get_inmate_info`) 6. Metrics comparison (optional: run `radon cc` or `pylint` on both versions) 7. Conclusion & lessons learned ### Optional: generate metrics ```bash pip install radon pylint radon cc legacy/prison_system.py -a radon cc -a refactored/prison_app ``` This gives you cyclomatic complexity numbers to compare in your report. # prison-management