canvas_backup

#!/usr/bin/python3

from lxml import html

import argparse
import http.cookiejar
import json
import os
import re
import requests
import string
import time

API_URL = '/api/v1/'
BASE_URL = None
COOKIE_FILE = None

ONE_MB = 2**20
ONE_KB = 2**10

FILE_COUNT = 0
TOTAL_SIZE = 0

def get_valid_filename(s):
    s = str(s).strip().replace(' ', '_')
    return re.sub(r'(?u)[^-\w.]', '', s)

def get_file_size(size):
    """
    Returns the file size in a human readable format. (e.g. 700.79 KB, 19.55 MB)
    """
    if size >= ONE_MB:
        return '{:.2f} MB'.format(size / ONE_MB)
    elif size >= ONE_KB:
        return '{:.2f} KB'.format(size / ONE_KB)
    return '{} {}'.format(size, 'bytes')

def write_file(url, filename, path, size, depth):
    """
    Writes a file to disk.
    """
    global TOTAL_SIZE
    global FILE_COUNT
    TOTAL_SIZE += size
    FILE_COUNT += 1
    readable_size = ''
    if size != 0:
        readable_size = get_file_size(size)

    print("{0:>9} {1}{2}".format(readable_size, '|-- ' * depth, filename))

    file = requests.get(url, cookies=COOKIE_FILE, stream=True)
    filename = get_valid_filename(filename)
    with open(path + '/' + filename, 'wb') as f:
        for block in file.iter_content(1024):
            f.write(block)

def download_file(file, path, depth):
    """
    Downloads a file given a canvas file object.
    """
    if not file or 'url' not in file or not file['url']:
        return 0

    write_file(file['url'], file['display_name'], path, file['size'], depth)
    return 0

def download_page(html_text, path, name, depth):
    """
    Scrapes a standard page for all files on it and downloads them.
    """
    files = html_text.find_class('instructure_file_link')
    if files:
        print("{0:9} {1}{2}/".format('', '|-- ' * depth, name))
        try:
            path = path + '/' + name
            os.mkdir(path)
        except:
            print('Failed to create an output directory.')
            return 1
    else:
        return 0

    for f in files:
        file_id = re.search('\/courses\/[0-9]*\/files\/([0-9]*)\/', f.get('href')).group(1)
        file_api = API_URL + 'files/' +  file_id
        file_response = json.loads(requests.get(file_api, cookies=COOKIE_FILE).text[9:])

        download_file(file_response, path, depth + 1)
    return 0

def download_quiz(quiz, path, depth):
    if not quiz or 'html_url' not in quiz or not quiz['html_url']:
        return 0

    name = get_valid_filename(quiz['title'])
    request = quiz['html_url']
    response = requests.get(request, cookies=COOKIE_FILE)

    if not response:
        return 0

    download_page(html.document_fromstring(response.content), path, name, depth)
    return 0

def download_module_body(page, path, depth):
    """
    Parses the body text of a module and finds files to download.
    """
    if not page or 'body' not in page or not page['body']:
        return 0

    name = get_valid_filename(page['title'])

    download_page(html.document_fromstring(page['body']), path, name, depth)
    return 0

def download_module(module, path, depth):
    """
    Downloads all the files in a given canvas module object.
    """
    if 'items_url' not in module or 'name' not in module:
        return 0

    api_request = module['items_url'] + '?per_page=200'
    name = get_valid_filename(module['name'])

    response = json.loads(requests.get(api_request, cookies=COOKIE_FILE).text[9:])
    if not response:
        return 0

    print("{0:9} {1}{2}/".format('', '|-- ' * depth, name))
    try:
        path = path + '/' + name
        os.mkdir(path)
    except:
        print('Failed to create an output directory.')
        return 1

    for r in response:
        if 'url' in r:
            object = json.loads(requests.get(r['url'], cookies=COOKIE_FILE).text[9:])
            if 'body' in object:
                # Object is a page: scrape the page for files
                download_module_body(object, path, depth + 1)
            elif 'quiz_type' in object:
                # Object is a quiz: scrape the quiz for files
                download_quiz(object, path, depth + 1)
            else:
                # Object is a file: download the file
                download_file(object, path, depth + 1)
    return 0

def download_modules(course_id, path, depth=1):
    """
    Recursively downloads all the modules in a course.
    """
    api_request = API_URL + 'courses/' + course_id + '/modules?per_page=200'
    response = json.loads(requests.get(api_request, cookies=COOKIE_FILE).text[9:])
    if not response:
        return 0

    print("{0:9} {1}Modules/".format('', '|-- ' * depth))
    try:
        os.mkdir(path)
    except:
        print('Failed to create an output directory.')
        return 1

    for r in response:
        download_module(r, path, depth + 1)
    return 0

def download_assignment(assignment_url, path, depth):
    """
    Downloads all the files on a given assignment.
    """
    response = requests.get(assignment_url, cookies=COOKIE_FILE)
    if not response:
        return 0

    response = html.document_fromstring(response.content)
    name = get_valid_filename(response.find_class('assignment-title')[0].find_class('title')[0].text)
    download_page(response, path, name, depth)
    return 0

def download_assignments(course_id, path, depth=1):
    """
    Recursively downloads all assignments for a course given a canvas course id.
    """
    api_request = API_URL + 'courses/' + course_id + '/students/submissions?per_page=200'
    response = json.loads(requests.get(api_request, cookies=COOKIE_FILE).text[9:])
    if not response:
        return 0

    print("{0:9} {1}Assignments/".format('', '|-- ' * depth))
    try:
        os.mkdir(path)
    except:
        print('Failed to create an output directory.')
        return 1

    for r in response:
        assignment_url = BASE_URL + '/courses/%s/assignments/%s' % (course_id, r['assignment_id'])
        download_assignment(assignment_url, path, depth+1)
    return 0

def download_files(api_request, path, depth):
    """
    Downloads all files in a folder given a file list canvas api request.
    """
    if not api_request:
        return 0

    response = json.loads(requests.get(api_request, cookies=COOKIE_FILE).text[9:])

    for r in response:
        download_file(r, path, depth)
    return 0

def download_folder(api_response, path, depth=1, root=False):
    """
    Recursively downloads a directory tree given a canvas folder object.
    """
    if not api_response or 'folders_url' not in api_response:
        return 0

    if root:
        print("{0:9} {1}Files/".format('', '|-- ' * depth))
        try:
            os.mkdir(path)
        except:
            print('Failed to create an output directory.')
            return 1
    else:
        print("{0:9} {1}{2}/".format('', '|-- ' * depth, api_response['name']))
        path = path + '/' + get_valid_filename(api_response['name'])
        try:
            os.mkdir(path)
        except:
            print('Failed to create an output directory.')
            return 1

    response = json.loads(requests.get(api_response['folders_url'], cookies=COOKIE_FILE).text[9:])
    download_files(api_response['files_url'], path, depth + 1)

    for r in response:
        download_folder(r, path, depth + 1)
    return 0

def download_root(name, asset_string, path):
    """
    Recursively downloads all content for a course.
    """
    print("{0:9} {1}/".format('', name))

    path = path + '/' + get_valid_filename(name)
    try:
        os.mkdir(path)
    except:
        print('Failed to create an output directory.')
        return 1

    asset_attr = asset_string.split('_')
    api_req = API_URL + asset_attr[0] + 's/' + asset_attr[1] + '/' + 'folders/root?per_page=200'
    response = json.loads(requests.get(api_req, cookies=COOKIE_FILE).text[9:])

    download_folder(response, path + '/Files', root=True)

    if asset_attr[0] == 'course':
        download_assignments(asset_attr[1], path + '/Assignments')
        download_modules(asset_attr[1], path + '/Modules')
    return 0

def download(args):
    global COOKIE_FILE
    COOKIE_FILE = http.cookiejar.MozillaCookieJar(args.cookies)
    try:
        COOKIE_FILE.load(ignore_expires=True)
    except:
        print('Failed to load the cookie file.\nMake sure the name of the file was entered correctly and that the first line of the cookie file is \'# HTTP Cookie File\'.')
        return 1

    for c in COOKIE_FILE:
        c.expires = time.time() + (24 * 3600)

    global API_URL
    global BASE_URL
    BASE_URL = args.url
    API_URL = args.url + API_URL

    response = requests.get(BASE_URL + '/files', cookies=COOKIE_FILE)
    if response.url != BASE_URL + '/files':
        print('Failed to access your canvas files.')
        return 1

    path = args.output
    try:
        os.mkdir(path)
    except FileExistsError:
        print('The directory %s already exists.' % path)
        return 1
    except:
        print('Failed to create an output directory.')
        return 1

    # Scrapes files page for courses, groups, and user's files.
    # Messier than making API calls for courses and groups but easier.
    folders = json.loads(html.document_fromstring(response.content).get_element_by_id('application').findtext('script').split(';')[1][9:])['FILES_CONTEXTS']
    courses = {}
    for f in folders:
        courses[f['asset_string']] = f['name']

    # Scrape courses page for missing courses since
    # the API doesn't return 'inactive' courses.
    courses_response = requests.get(BASE_URL + '/courses', cookies=COOKIE_FILE)
    courses_response = html.document_fromstring(courses_response.content).find_class('course-list-table-row')
    for c in courses_response:
        a = c.find_class('course-list-course-title-column')[0].find('a')
        if a is not None:
            asset_string = 'course_' + a.get('href').split('/')[2]
            name = a.get('title')
            if asset_string not in courses:
                courses[asset_string] = name

    for asset_string, name in courses.items():
        download_root(name, asset_string, path)
        print('')

    print('%s files downloaded (%s)' % (FILE_COUNT, get_file_size(TOTAL_SIZE)))
    return 0

def main(argv=None):
    parser = argparse.ArgumentParser(description='Downloads a complete backup of bCourses.')
    parser.usage = """
    You need to supply a cookie.txt from your bCourses session.
    This can be done, for example, by using the cookies.txt extension in Google Chrome.
    You can then grab your cookies.txt file from your canvas page (e.g. https://bcourses.berkeley.edu)."""

    parser.add_argument('--cookies', type=str, default='cookies.txt', help='Path to the cookiefile.')
    parser.add_argument('--output', type=str, default='downloads', help='Path to the output directory.')
    parser.add_argument('--url', type=str, default='https://bcourses.berkeley.edu', help='URL to canvas website.')
    args = parser.parse_args(argv)

    return download(args)

if __name__ == '__main__':
    main()