main.py

import subprocess
import os
import time
import sys
import urllib.request
import urllib.error

# The server for cloning student files. For example, you usually clone from
# the server "https://github.com"
git_server = "https://github.com"

if len(sys.argv) is not 3:
    print("usage: python {} project-name users-file.txt".format(sys.argv[0]))
    exit(1)

project_name = sys.argv[1]
users_file = sys.argv[2]

with open(users_file) as f:
    users = f.read().splitlines()

# Create the reports directories
os.makedirs(os.path.join("reports", project_name, "good"), exist_ok=True)
os.makedirs(os.path.join("reports", project_name, "bad"), exist_ok=True)

for user in users:
    # If this is an http git server
    if 'http' in git_server:
        # Then confirm that the repo exists before trying to clone it
        try:
            response = urllib.request.urlopen("/".join([git_server, user, project_name]))
        except urllib.error.HTTPError:
            continue

    # The path to this user's homework submission
    repo_path = os.path.join('users', user, project_name)

    # Remove any existing submission that we've downloaded; we're going to test
    # whatever is in the repo right now.
    subprocess.call(
        [
            "rm",
            "-rf",
            repo_path,
        ]
    )

    # Clone down the user's submission into repo_path
    result = subprocess.call(
        [
            "git",
            "clone",
            "{}/{}/{}".format(git_server, user, project_name),
            repo_path,
        ]
    )

    subprocess.call(
        [
            "cp",
            os.path.join("tests", project_name, "tests.py"),
            repo_path,
        ]
    )

    # Did we successfully clone down the project?
    clone_successful = result is 0

    # If not, then skip this student
    if not clone_successful:
        continue

    time.sleep(3)

    # Run the tests, collecting stdout and stderr
    stdout = b""
    stderr = b""

    test_process = subprocess.Popen(
        [
            "python",
            #os.path.join("tests", "tests.py"),
            "tests.py"
        ],
        cwd=repo_path,
        stdout=subprocess.PIPE,
        stderr=subprocess.PIPE,
    )

    try:
        # Hope for the best
        success = True

        try:
            # Retrieve stdout and stderr
            stdout, stderr = test_process.communicate(timeout=10)
        except subprocess.TimeoutExpired:
            # If the tests took too long, then kill the process, record tests as failed
            test_process.kill()
            success = False

        if test_process.returncode is None or test_process.returncode is not 0:
            # If the tests failed, mark them as failed
            success = False

        # Build the paths for this student's good/bad reports
        good_result_path = os.path.join("reports", project_name, "good", "{}.txt".format(user))
        bad_result_path = os.path.join("reports", project_name, "bad", "{}.txt".format(user))

        # Reformat stderr and stdout
        contents = (stderr.decode().replace("\r\n", "\n") + "\n\n" + stdout.decode().replace("\r\n", "\n"))

        # Remove any existing good/bad reports
        for file_path in [bad_result_path, good_result_path]:
            try:
                os.unlink(file_path)
            except FileNotFoundError:
                pass

        if success:
            result_path = good_result_path
        else:
            result_path = bad_result_path

        with open(result_path, "w") as f:
            f.write(contents)

    finally:
        # In any case, kill the test process
        test_process.kill()