CodeEval

Build

The application could be build with Docker

First navigate to the project root directory and build the image:

$ docker build -t code_eval:latest .

After build completes, we can run the container:

$ docker run -d -p 5000:5000 code_eval

Sample Login Credentials

Default Admin

Email: example_admin_user@gmail.com
Password: 111

Default User

Email: example_user@gmail.com
Password: 111

Functionality (Student)

Join Course

To join a course, students need to first visit the link localhost:5000/register/<course_invitation> to gain access for code submissions.

(By default visit localhost:5000/register/join156 to add CS156.)

Submit Code

To submit a valid code, students need to create an entry function for tester to run (by default the entry function is named 'entry') and return the correct result according to instructions.

(You could refer to this sample file which does a simple add function.)

Functionality (Professor)

Submit Code

< Same as Above >

Summary

Professors can view a summary of student submissions to each course and session.

The summary is structured as Index >> Course >> Session >> Student >> Submissions >> Result .

Upload Session

Upload a new session to a specific course.

Required parameters:

• Filename: A file consisting of all test cases.

Example:

# import BaseTest is necessary for integration
from base import BaseTest

class TestCases(BaseTest):

    def __init__(self, func):
        super().__init__(func)
        
        # each key represent a separate question
        self.parameters = {
            'add': [(1, 2), (3, 4), (7, 10), (5, 6)],
            'subtract': [(1, 2), (3, 4), (7, 10), (5, 6)],
            'multiply': [(1, 2), (3, 4), (7, 10), (5, 6)]
        }
        self.answers = {
            'add': [3, 7, 17, 11],
            'subtract': [-1, -1, -3, -1],
            'multiply': [2, 12, 70, 30]
        }

• Session Number: Should be a positive decimal (e.g. 1.1).
• Course: Choose from available courses.
• Blacklist: Libraries that students should not use in submissions.

Add Course

Add a new course and create a registration link for that course.

All Settings

View / Modify / Delete courses and sessions.

Unit Test

$ python3 -m unittest

Plagiarism Detection

The Autograder comes with a source code plagiarism detection module that applies the following algorithms:

• Exact Match
• Unifying AST
• Unifying AST (Ignore Variables)
• Winnowing
• Tree Edit Distance (RTED)
• Comment Edit Distance (Levenshtein)

A CLI tool is included in the repo for experimenting with the above algorithms on different attack cases. To use the tool, navigate to the plagiarism directory and run the following command:

python3 -m plagiarism.py