Code Judge (FastAPI Backend)

A lightweight Online Code Judge built using Flask, designed to replicate the core functionality of platforms like Codeforces or LeetCode. It executes user-submitted Python code, evaluates it against predefined test cases, and provides detailed verdicts. This project serves as a foundational implementation to understand the intricacies of automated code evaluation systems, offering a pocket-sized, yet powerful, solution for competitive programming practice or educational purposes.

✨ Key Features

✅ Execute Python code submissionsSecurely runs user-provided Python scripts in an isolated environment.
✅ Accept custom input (stdin)Allows test cases to be dynamically supplied to the executing code via standard input.
✅ Compare output with expected outputAutomatically verifies the correctness of the submitted code's output against predefined correct solutions.
✅ Comprehensive Verdict SystemProvides clear feedback on submission results, including Accepted, Wrong Answer, Runtime Error, and Time Limit Exceeded.
✅ Execution time limitsEnforces strict time constraints on code execution to prevent infinite loops and inefficient solutions.
✅ Clean APIOffers a well-structured and easy-to-use API endpoint for submitting code and retrieving results programmatically.
✅ JSON-based request/responseUtilizes industry-standard JSON format for seamless data exchange between clients and the judge server.
✅ Beginner-friendly but system-level logicDesigned with an approachable architecture, while incorporating fundamental system-level concepts essential for robust code execution and evaluation.

🧠 How It Works

1.
User sends Python code via API: Submissions are made through a dedicated REST API endpoint.
2.
Code is written to a temporary file: The received code is saved to a secure, temporary location on the server filesystem.
3.
Code is executed using `subprocess`: Python's `subprocess` module is leveraged to run the user's code as a separate process.
4.
Input is piped through stdin: Test case inputs are redirected to the executing code's standard input stream.
5.
Output is captured from stdout: The executed code's standard output is captured for later comparison.
6.
Output is compared with expected output: The captured output is then meticulously compared against the known correct output.
7.
Judge returns a detailed verdict: Based on the comparison and execution metrics, a final verdict is determined.
⚠️ Note on Sandboxing: Current implementation provides basic process isolation primarily through timeouts. Advanced security measures like Docker containerization are considered for future enhancements.

🛠 Tech Stack

🐍
PythonThe core programming language for the backend logic and code execution.
🌶
FastAPIA lightweight web framework used to build the API for handling submissions.
⚙️
subprocess moduleEssential for spawning and managing external processes to execute user code.
📄
tempfile moduleUtilized for securely creating and managing temporary files for code storage and execution.
🧪
PostmanUsed extensively for API development, testing, and debugging the backend endpoints.

Understanding the Verdict System

🟢

AC — Accepted

Correct output for all test cases within limits.

🔴

WA — Wrong Answer

Output did not match expected output.

⚠️

RE — Runtime Error

Code terminated abnormally (crash/exception).

⏱

TLE — Time Limit

Execution exceeded maximum allowed time.