#!/usr/bin/env python from __future__ import print_function import argparse import base64 import calendar import codecs import errno import getpass import json import logging import os import re import select import subprocess import sys import time import urlparse import urllib import urllib2 from github_backup import __version__ FNULL = open(os.devnull, 'w') def log_error(message): if type(message) == str: message = [message] for msg in message: sys.stderr.write("{0}\n".format(msg)) sys.exit(1) def log_info(message): if type(message) == str: message = [message] for msg in message: sys.stdout.write("{0}\n".format(msg)) def logging_subprocess(popenargs, logger, stdout_log_level=logging.DEBUG, stderr_log_level=logging.ERROR, **kwargs): """ Variant of subprocess.call that accepts a logger instead of stdout/stderr, and logs stdout messages via logger.debug and stderr messages via logger.error. """ child = subprocess.Popen(popenargs, stdout=subprocess.PIPE, stderr=subprocess.PIPE, **kwargs) log_level = {child.stdout: stdout_log_level, child.stderr: stderr_log_level} def check_io(): ready_to_read = select.select([child.stdout, child.stderr], [], [], 1000)[0] for io in ready_to_read: line = io.readline() if not logger: continue if not (io == child.stderr and not line): logger.log(log_level[io], line[:-1]) # keep checking stdout/stderr until the child exits while child.poll() is None: check_io() check_io() # check again to catch anything after the process exits rc = child.wait() if rc != 0: print(u'{} returned {}:'.format(popenargs[0], rc), file=sys.stderr) print('\t', u' '.join(popenargs), file=sys.stderr) return rc def mkdir_p(*args): for path in args: try: os.makedirs(path) except OSError as exc: # Python >2.5 if exc.errno == errno.EEXIST and os.path.isdir(path): pass else: raise def mask_password(url, secret='*****'): parsed = urlparse.urlparse(url) if not parsed.password: return url elif parsed.password == 'x-oauth-basic': return url.replace(parsed.username, secret) return url.replace(parsed.password, secret) def parse_args(): parser = argparse.ArgumentParser(description='Backup a github account', prog='Github Backup') parser.add_argument('user', metavar='USER', type=str, help='github username') parser.add_argument('-u', '--username', dest='username', help='username for basic auth') parser.add_argument('-p', '--password', dest='password', help='password for basic auth. ' 'If a username is given but not a password, the ' 'password will be prompted for.') parser.add_argument('-t', '--token', dest='token', help='personal access or OAuth token') parser.add_argument('-o', '--output-directory', default='.', dest='output_directory', help='directory at which to backup the repositories') parser.add_argument('-i', '--incremental', action='store_true', dest='incremental', help='incremental backup') parser.add_argument('--starred', action='store_true', dest='include_starred', help='include starred repositories in backup') parser.add_argument('--watched', action='store_true', dest='include_watched', help='include watched repositories in backup') parser.add_argument('--all', action='store_true', dest='include_everything', help='include everything in backup') parser.add_argument('--issues', action='store_true', dest='include_issues', help='include issues in backup') parser.add_argument('--issue-comments', action='store_true', dest='include_issue_comments', help='include issue comments in backup') parser.add_argument('--issue-events', action='store_true', dest='include_issue_events', help='include issue events in backup') parser.add_argument('--pulls', action='store_true', dest='include_pulls', help='include pull requests in backup') parser.add_argument('--pull-comments', action='store_true', dest='include_pull_comments', help='include pull request review comments in backup') parser.add_argument('--pull-commits', action='store_true', dest='include_pull_commits', help='include pull request commits in backup') parser.add_argument('--labels', action='store_true', dest='include_labels', help='include labels in backup') parser.add_argument('--hooks', action='store_true', dest='include_hooks', help='include hooks in backup (works only when authenticated)') parser.add_argument('--milestones', action='store_true', dest='include_milestones', help='include milestones in backup') parser.add_argument('--repositories', action='store_true', dest='include_repository', help='include repository clone in backup') parser.add_argument('--wikis', action='store_true', dest='include_wiki', help='include wiki clone in backup') parser.add_argument('--skip-existing', action='store_true', dest='skip_existing', help='skip project if a backup directory exists') parser.add_argument('-L', '--languages', dest='languages', help='only allow these languages', nargs='*') parser.add_argument('-N', '--name-regex', dest='name_regex', help='python regex to match names against') parser.add_argument('-H', '--github-host', dest='github_host', help='GitHub Enterprise hostname') parser.add_argument('-O', '--organization', action='store_true', dest='organization', help='whether or not this is an organization user') parser.add_argument('-R', '--repository', dest='repository', help='name of repository to limit backup to') parser.add_argument('-P', '--private', action='store_true', dest='private', help='include private repositories') parser.add_argument('-F', '--fork', action='store_true', dest='fork', help='include forked repositories') parser.add_argument('--prefer-ssh', action='store_true', help='Clone repositories using SSH instead of HTTPS') parser.add_argument('-v', '--version', action='version', version='%(prog)s ' + __version__) return parser.parse_args() def get_auth(args, encode=True): auth = None if args.token: auth = args.token + ':' + 'x-oauth-basic' elif args.username: if not args.password: args.password = getpass.getpass() auth = args.username + ':' + args.password elif args.password: log_error('You must specify a username for basic auth') if not auth: return None if encode == False: return auth return base64.b64encode(auth) def get_github_api_host(args): if args.github_host: host = args.github_host + '/api/v3' else: host = 'api.github.com' return host def get_github_host(args): if args.github_host: host = args.github_host else: host = 'github.com' return host def get_github_repo_url(args, repository): if args.prefer_ssh: return repository['ssh_url'] auth = get_auth(args, False) if auth: repo_url = 'https://{0}@{1}/{2}/{3}.git'.format( auth, get_github_host(args), args.user, repository['name']) else: repo_url = repository['clone_url'] return repo_url def retrieve_data(args, template, query_args=None, single_request=False): auth = get_auth(args) query_args = get_query_args(query_args) per_page = 100 page = 0 data = [] while True: page = page + 1 request = _construct_request(per_page, page, query_args, template, auth) # noqa r, errors = _get_response(request, auth, template) status_code = int(r.getcode()) if status_code != 200: template = 'API request returned HTTP {0}: {1}' errors.append(template.format(status_code, r.reason)) log_error(errors) response = json.loads(r.read()) if len(errors) == 0: if type(response) == list: data.extend(response) if len(response) < per_page: break elif type(response) == dict and single_request: data.append(response) if len(errors) > 0: log_error(errors) if single_request: break return data def get_query_args(query_args=None): if not query_args: query_args = {} return query_args def _get_response(request, auth, template): retry_timeout = 3 errors = [] # We'll make requests in a loop so we can # delay and retry in the case of rate-limiting while True: should_continue = False try: r = urllib2.urlopen(request) except urllib2.HTTPError as exc: errors, should_continue = _request_http_error(exc, auth, errors) # noqa r = exc except urllib2.URLError: should_continue = _request_url_error(template, retry_timeout) if not should_continue: raise if should_continue: continue break return r, errors def _construct_request(per_page, page, query_args, template, auth): querystring = urllib.urlencode(dict({ 'per_page': per_page, 'page': page }.items() + query_args.items())) request = urllib2.Request(template + '?' + querystring) if auth is not None: request.add_header('Authorization', 'Basic ' + auth) return request def _request_http_error(exc, auth, errors): # HTTPError behaves like a Response so we can # check the status code and headers to see exactly # what failed. should_continue = False headers = exc.headers limit_remaining = int(headers.get('x-ratelimit-remaining', 0)) if exc.code == 403 and limit_remaining < 1: # The X-RateLimit-Reset header includes a # timestamp telling us when the limit will reset # so we can calculate how long to wait rather # than inefficiently polling: gm_now = calendar.timegm(time.gmtime()) reset = int(headers.get('x-ratelimit-reset', 0)) or gm_now # We'll never sleep for less than 10 seconds: delta = max(10, reset - gm_now) limit = headers.get('x-ratelimit-limit') print('Exceeded rate limit of {} requests; waiting {} seconds to reset'.format(limit, delta), # noqa file=sys.stderr) ratelimit_error = 'No more requests remaining' if auth is None: ratelimit_error += '; authenticate to raise your GitHub rate limit' # noqa errors.append(ratelimit_error) time.sleep(delta) should_continue = True return errors, should_continue def _request_url_error(template, retry_timeout): # Incase of a connection timing out, we can retry a few time # But we won't crash and not back-up the rest now log_info('{} timed out'.format(template)) retry_timeout -= 1 if retry_timeout >= 0: return True log_error('{} timed out to much, skipping!') return False def retrieve_repositories(args): log_info('Retrieving repositories') single_request = False template = 'https://{0}/user/repos'.format( get_github_api_host(args)) if args.organization: template = 'https://{0}/orgs/{1}/repos'.format( get_github_api_host(args), args.user) if args.repository: single_request = True template = 'https://{0}/repos/{1}/{2}'.format( get_github_api_host(args), args.user, args.repository) return retrieve_data(args, template, single_request=single_request) def filter_repositories(args, repositories): log_info('Filtering repositories') repositories = [r for r in repositories if r['owner']['login'] == args.user] name_regex = None if args.name_regex: name_regex = re.compile(args.name_regex) languages = None if args.languages: languages = [x.lower() for x in args.languages] if not args.fork: repositories = [r for r in repositories if not r['fork']] if not args.private: repositories = [r for r in repositories if not r['private']] if languages: repositories = [r for r in repositories if r['language'] and r['language'].lower() in languages] # noqa if name_regex: repositories = [r for r in repositories if name_regex.match(r['name'])] return repositories def backup_repositories(args, output_directory, repositories): log_info('Backing up repositories') repos_template = 'https://{0}/repos'.format(get_github_api_host(args)) if args.incremental: last_update = max(repository['updated_at'] for repository in repositories) last_update_path = os.path.join(output_directory, 'last_update') if os.path.exists(last_update_path): args.since = open(last_update_path).read().strip() else: args.since = None else: args.since = None for repository in repositories: backup_cwd = os.path.join(output_directory, 'repositories') repo_cwd = os.path.join(backup_cwd, repository['name']) repo_dir = os.path.join(repo_cwd, 'repository') repo_url = get_github_repo_url(args, repository) if args.include_repository or args.include_everything: fetch_repository(repository['name'], repo_url, repo_dir, skip_existing=args.skip_existing) download_wiki = (args.include_wiki or args.include_everything) if repository['has_wiki'] and download_wiki: fetch_repository(repository['name'], repo_url.replace('.git', '.wiki.git'), os.path.join(repo_cwd, 'wiki'), skip_existing=args.skip_existing) if args.include_issues or args.include_everything: backup_issues(args, repo_cwd, repository, repos_template) if args.include_pulls or args.include_everything: backup_pulls(args, repo_cwd, repository, repos_template) if args.include_milestones or args.include_everything: backup_milestones(args, repo_cwd, repository, repos_template) if args.include_labels or args.include_everything: backup_labels(args, repo_cwd, repository, repos_template) if args.include_hooks or args.include_everything: backup_hooks(args, repo_cwd, repository, repos_template) if args.incremental: open(last_update_path, 'w').write(last_update) def backup_issues(args, repo_cwd, repository, repos_template): has_issues_dir = os.path.isdir('{0}/issues/.git'.format(repo_cwd)) if args.skip_existing and has_issues_dir: return log_info('Retrieving {0} issues'.format(repository['full_name'])) issue_cwd = os.path.join(repo_cwd, 'issues') mkdir_p(repo_cwd, issue_cwd) issues = {} issues_skipped = 0 issues_skipped_message = '' _issue_template = '{0}/{1}/issues'.format(repos_template, repository['full_name']) issue_states = ['open', 'closed'] for issue_state in issue_states: query_args = { 'filter': 'all', 'state': issue_state } if args.since: query_args['since'] = args.since _issues = retrieve_data(args, _issue_template, query_args=query_args) for issue in _issues: # skip pull requests which are also returned as issues # if retrieving pull requests is requested as well if 'pull_request' in issue and (args.include_pulls or args.include_everything): issues_skipped += 1 continue issues[issue['number']] = issue if issues_skipped: issues_skipped_message = ' (skipped {0} pull requests)'.format(issues_skipped) log_info('Saving {0} issues to disk{1}'.format(len(issues.keys()), issues_skipped_message)) comments_template = _issue_template + '/{0}/comments' events_template = _issue_template + '/{0}/events' for number, issue in issues.iteritems(): if args.include_issue_comments or args.include_everything: template = comments_template.format(number) issues[number]['comment_data'] = retrieve_data(args, template) if args.include_issue_events or args.include_everything: template = events_template.format(number) issues[number]['event_data'] = retrieve_data(args, template) issue_file = '{0}/{1}.json'.format(issue_cwd, number) with codecs.open(issue_file, 'w', encoding='utf-8') as f: json_dump(issue, f) def backup_pulls(args, repo_cwd, repository, repos_template): has_pulls_dir = os.path.isdir('{0}/pulls/.git'.format(repo_cwd)) if args.skip_existing and has_pulls_dir: return log_info('Retrieving {0} pull requests'.format(repository['full_name'])) # noqa pulls_cwd = os.path.join(repo_cwd, 'pulls') mkdir_p(repo_cwd, pulls_cwd) pulls = {} _pulls_template = '{0}/{1}/pulls'.format(repos_template, repository['full_name']) pull_states = ['open', 'closed'] for pull_state in pull_states: query_args = { 'filter': 'all', 'state': pull_state, 'sort': 'updated', 'direction': 'desc', } # It'd be nice to be able to apply the args.since filter here... _pulls = retrieve_data(args, _pulls_template, query_args=query_args) for pull in _pulls: if not args.since or pull['updated_at'] >= args.since: pulls[pull['number']] = pull log_info('Saving {0} pull requests to disk'.format(len(pulls.keys()))) comments_template = _pulls_template + '/{0}/comments' commits_template = _pulls_template + '/{0}/commits' for number, pull in pulls.iteritems(): if args.include_pull_comments or args.include_everything: template = comments_template.format(number) pulls[number]['comment_data'] = retrieve_data(args, template) if args.include_pull_commits or args.include_everything: template = commits_template.format(number) pulls[number]['commit_data'] = retrieve_data(args, template) pull_file = '{0}/{1}.json'.format(pulls_cwd, number) with codecs.open(pull_file, 'w', encoding='utf-8') as f: json_dump(pull, f) def backup_milestones(args, repo_cwd, repository, repos_template): milestone_cwd = os.path.join(repo_cwd, 'milestones') if args.skip_existing and os.path.isdir(milestone_cwd): return log_info('Retrieving {0} milestones'.format(repository['full_name'])) mkdir_p(repo_cwd, milestone_cwd) template = '{0}/{1}/milestones'.format(repos_template, repository['full_name']) query_args = { 'state': 'all' } _milestones = retrieve_data(args, template, query_args=query_args) milestones = {} for milestone in _milestones: milestones[milestone['number']] = milestone log_info('Saving {0} milestones to disk'.format(len(milestones.keys()))) for number, milestone in milestones.iteritems(): milestone_file = '{0}/{1}.json'.format(milestone_cwd, number) with codecs.open(milestone_file, 'w', encoding='utf-8') as f: json_dump(milestone, f) def backup_labels(args, repo_cwd, repository, repos_template): label_cwd = os.path.join(repo_cwd, 'labels') output_file = '{0}/labels.json'.format(label_cwd) template = '{0}/{1}/labels'.format(repos_template, repository['full_name']) _backup_data(args, 'labels', template, output_file, label_cwd) def backup_hooks(args, repo_cwd, repository, repos_template): auth = get_auth(args) if not auth: log_info("Skipping hooks since no authentication provided") return hook_cwd = os.path.join(repo_cwd, 'hooks') output_file = '{0}/hooks.json'.format(hook_cwd) template = '{0}/{1}/hooks'.format(repos_template, repository['full_name']) try: _backup_data(args, 'hooks', template, output_file, hook_cwd) except SystemExit: log_info("Unable to read hooks, skipping") def fetch_repository(name, remote_url, local_dir, skip_existing=False): clone_exists = os.path.exists(os.path.join(local_dir, '.git')) if clone_exists and skip_existing: return masked_remote_url = mask_password(remote_url) initalized = subprocess.call('git ls-remote ' + remote_url, stdout=FNULL, stderr=FNULL, shell=True) if initalized == 128: log_info("Skipping {0} ({1}) since it's not initalized".format(name, masked_remote_url)) return if clone_exists: log_info('Updating {0} in {1}'.format(name, local_dir)) git_command = ['git', 'remote', 'rm', 'origin'] logging_subprocess(git_command, None, cwd=local_dir) git_command = ['git', 'remote', 'add', 'origin', remote_url] logging_subprocess(git_command, None, cwd=local_dir) git_command = ['git', 'fetch', '--all', '--tags', '--prune'] logging_subprocess(git_command, None, cwd=local_dir) else: log_info('Cloning {0} repository from {1} to {2}'.format(name, masked_remote_url, local_dir)) git_command = ['git', 'clone', remote_url, local_dir] logging_subprocess(git_command, None) def backup_account(args, output_directory): account_cwd = os.path.join(output_directory, 'account') if args.include_starred or args.include_everything: output_file = '{0}/starred.json'.format(account_cwd) template = "https://{0}/users/{1}/starred" template = template.format(get_github_api_host(args), args.user) _backup_data(args, 'starred repositories', template, output_file, account_cwd) if args.include_watched or args.include_everything: output_file = '{0}/watched.json'.format(account_cwd) template = "https://{0}/users/{1}/subscriptions" template = template.format(get_github_api_host(args), args.user) _backup_data(args, 'watched repositories', template, output_file, account_cwd) def _backup_data(args, name, template, output_file, output_directory): skip_existing = args.skip_existing if not skip_existing or not os.path.exists(output_file): log_info('Retrieving {0} {1}'.format(args.user, name)) mkdir_p(output_directory) data = retrieve_data(args, template) log_info('Writing {0} {1} to disk'.format(len(data), name)) with codecs.open(output_file, 'w', encoding='utf-8') as f: json_dump(data, f) def json_dump(data, output_file): json.dump(data, output_file, ensure_ascii=False, sort_keys=True, indent=4, separators=(',', ': ')) def main(): args = parse_args() output_directory = os.path.realpath(args.output_directory) if not os.path.isdir(output_directory): log_info('Create output directory {0}'.format(output_directory)) mkdir_p(output_directory) log_info('Backing up user {0} to {1}'.format(args.user, output_directory)) repositories = retrieve_repositories(args) repositories = filter_repositories(args, repositories) backup_repositories(args, output_directory, repositories) backup_account(args, output_directory) if __name__ == '__main__': main()