Release version 0.10.1

Implement incremental updates

CHANGES.rst

@@ -1,6 +1,35 @@
 Changelog
 =========
 
+0.10.1 (2016-08-20)
+-------------------
+
+- Implement incremental updates. [Robert Bradshaw]
+
+  Guarded with an --incremental flag.
+
+  Stores the time of the last update and only downloads issue and
+  pull request data since this time.  All other data is relatively
+  small (likely fetched with a single request) and so is simply
+  re-populated from scratch as before.
+
+
+0.9.0 (2016-03-29)
+------------------
+
+- Fix cloning private repos with basic auth or token. [Kazuki Suda]
+
+0.8.0 (2016-02-14)
+------------------
+
+- Don't store issues which are actually pull requests. [Enrico Tröger]
+
+  This prevents storing pull requests twice since the Github API returns
+  pull requests also as issues. Those issues will be skipped but only if
+  retrieving pull requests is requested as well.
+  Closes #23.
+
+
 0.7.0 (2016-02-02)
 ------------------
 

bin/github-backup

@@ -16,6 +16,7 @@ import select
 import subprocess
 import sys
 import time
+import urlparse
 import urllib
 import urllib2
 
@@ -95,6 +96,15 @@ def mkdir_p(*args):
             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',
@@ -122,6 +132,11 @@ def parse_args():
                         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',
@@ -221,20 +236,26 @@ def parse_args():
     return parser.parse_args()
 
 
-def get_auth(args):
-    if args.token:
-        return base64.b64encode(args.token + ':' + 'x-oauth-basic')
+def get_auth(args, encode=True):
+    auth = None
 
-    if args.username:
+    if args.token:
+        auth = args.token + ':' + 'x-oauth-basic'
+    elif args.username:
         if not args.password:
             args.password = getpass.getpass()
-        return base64.b64encode(args.username + ':' + args.password)
-
-    if args.password:
+        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:
@@ -245,7 +266,7 @@ def get_github_api_host(args):
     return host
 
 
-def get_github_ssh_host(args):
+def get_github_host(args):
     if args.github_host:
         host = args.github_host
     else:
@@ -253,6 +274,21 @@ def get_github_ssh_host(args):
 
     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)
@@ -428,15 +464,21 @@ 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')
-
-        if args.prefer_ssh:
-            repo_url = repository['ssh_url']
-        else:
-            repo_url = repository['clone_url']
+        repo_url = get_github_repo_url(args, repository)
 
         if args.include_repository or args.include_everything:
             fetch_repository(repository['name'],
@@ -466,6 +508,8 @@ def backup_repositories(args, output_directory, repositories):
         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))
@@ -477,6 +521,8 @@ def backup_issues(args, repo_cwd, repository, repos_template):
     mkdir_p(repo_cwd, issue_cwd)
 
     issues = {}
+    issues_skipped = 0
+    issues_skipped_message = ''
     _issue_template = '{0}/{1}/issues'.format(repos_template,
                                               repository['full_name'])
 
@@ -486,14 +532,24 @@ def backup_issues(args, repo_cwd, repository, repos_template):
             '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
 
-    log_info('Saving {0} issues to disk'.format(len(issues.keys())))
+    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():
@@ -526,13 +582,17 @@ def backup_pulls(args, repo_cwd, repository, repos_template):
     for pull_state in pull_states:
         query_args = {
             'filter': 'all',
-            'state': pull_state
+            '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())))
@@ -616,12 +676,14 @@ def fetch_repository(name, remote_url, local_dir, skip_existing=False):
     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, remote_url))
+        log_info("Skipping {0} ({1}) since it's not initalized".format(name, masked_remote_url))
         return
 
     if clone_exists:
@@ -634,7 +696,7 @@ def fetch_repository(name, remote_url, local_dir, skip_existing=False):
         logging_subprocess(git_command, None, cwd=local_dir)
     else:
         log_info('Cloning {0} repository from {1} to {2}'.format(name,
-                                                                 remote_url,
+                                                                 masked_remote_url,
                                                                  local_dir))
         git_command = ['git', 'clone', remote_url, local_dir]
         logging_subprocess(git_command, None)

github_backup/__init__.py

@@ -1 +1 @@
-__version__ = '0.7.0'
+__version__ = '0.10.1'