examshell-python

Interactive terminal shell for practising 42-style Python exam tasks. Simulates the real exam environment — timed pressure, blind grading, sequential levels — or lets you work through exercises freely in practice mode.

Features

• Two modes — Real (exam simulation) and Practice (free training)
• Automatic grading — grademe runs your solution against test cases and shows a results table
• Multi-language task descriptions — each task is described in EN, DE, FR, ES, IT
• 14 tasks across 6 levels — from easy warm-ups to hard algorithmic challenges
• Clean terminal UI — built with Rich

Installation

Requires Python 3.10+.

pip install examshell-python

Or install from source:

git clone <repo-url>
cd examshell-python
pip install .

Usage

examshell42

On launch you choose an available exam rank, then a mode. In the mode menu, choose b to return to rank selection:

╭─ Choose your Exam mode ─╮
│ Real mode       (1)     │
│ Practice mode   (2)     │
│ Best solutions  (3)     │
│ Back to rank    (b)     │
╰─────────────────────────╯

At startup, examshell attempts to refresh ranks 03, 04, and 05 from the exam sites and saves valid task data as rank*.json. If an update fails, existing local task files remain available and the application continues offline.

Real mode

Simulates the actual exam. Tasks are drawn randomly per level. Progress is distributed across the available levels; completing the final level reaches 100.

Practice mode

Work through all tasks in any order, retry freely, no score pressure.

Best solutions

Downloads available reference solutions into the session's solutions/ folder without running them or overwriting existing files. The public source repository uses rankN/<task-name>/<task-file> paths; missing solutions are reported in the results table.

How it works

When a session starts, examshell creates a working directory (exam42/) with two sub-folders:

exam42/
├── subjects/
│   └── <task_name>/
│       ├── <task_name>.en.txt   ← task description + signature + examples
│       ├── <task_name>.de.txt
│       └── ...
└── rendu/
    └── <task_name>/
        └── <task_name>.py       ← put your solution here

Write your solution in rendu/<task_name>/<task_name>.py, then run grademe. The shell imports your file, executes each example against it, and prints a pass/fail table.

Tasks

Dependencies

• pyjsparser ≥ 2.7.1 — parse downloaded task data
• pydantic ≥ 2.0 — task data models
• rich ≥ 13.0 — terminal UI
• requests ≥ 2.0 — download task data

Project structure

src/examshell/
├── __init__.py
├── __main__.py
├── main.py           # CLI entry point, mode logic, UI screens
├── models.py         # Pydantic models (Task, Description, Example)
├── task_generator.py # TaskManager: file creation, grading, task selection
├── config.py
└── data/
    └── data.json     # All task definitions

License

MIT