move repository logic to its own module

Looking at update-committed-patches in the linaro-patchmetrics file and
seeing where this is headed, I think now is the time to split
functionality up so its easier to manage.

This also improves the logic of managing the last analyzed commit by
introducing a process_unchecked_commits method. The result of the change
also simplifies/improves some of the logic needed for unit-testing.

Change-Id: I1e28f3fe3b27e1afd414939d28acc5eb5f6387cf

1 files changed, 8 insertions, 49 deletions

diff --git a/update_commited_patches.py b/update_commited_patches.py
index a342028..5fba8a4 100755
--- a/update_commited_patches.py
+++ b/update_commited_patches.py
@@ -8,6 +8,8 @@ django_setup()  # must be called to get sys.path and django settings in place
 
 from patchwork.models import Patch, Project, State
 
+import gitrepo
+
 log = get_logger('update_commited_patches')
 
 _here = os.path.abspath(os.path.dirname(__file__))
@@ -18,45 +20,10 @@ def _assert_repo_dir(path):
         os.mkdir(path)
 
 
-def _clone_repo(repo_dir, giturl, linkname):
-    args = ['git', 'clone', giturl, linkname]
-    subprocess.check_call(args, cwd=repo_dir)
-
-
-def _pull(gitdir):
-    timeout = str(getattr(settings, 'REPO_TIMEOUT', 120))
-    subprocess.check_call(['timeout', timeout, 'git', 'fetch'], cwd=gitdir)
-    commit = subprocess.check_output(
-        ['git', 'merge-base', 'origin', 'master'], cwd=gitdir)
-    subprocess.check_call(
-        ['git', 'reset', '--hard', commit.strip()], cwd=gitdir)
-    subprocess.check_call(['timeout', timeout, 'git', 'pull'], cwd=gitdir)
-
-
-def _commits_to_check(gitdir):
-    start_at = os.path.join(gitdir, 'patchwork-last-commit')
-    if not os.path.exists(start_at):
-        start_at = 'HEAD~100'
-    else:
-        with open(start_at) as f:
-            start_at = f.read().strip()
-
-    args = ['git', 'rev-list', '--reverse', start_at + '..HEAD']
-    with open('/dev/null', 'w') as f:
-        rc = subprocess.call(
-            ['git', 'show', start_at], cwd=gitdir, stdout=f, stderr=f)
-    if rc != 0:
-        # we may have had a branch who's history was re-written
-        # just try and get changes for past day
-        args = ['git', 'rev-list', '--reverse', '--since', '1 day ago', 'HEAD']
-    commits = subprocess.check_output(args, cwd=gitdir).split('\n')
-    return [x for x in commits if len(x)]
-
-
-def _update_commit(project, gitdir, commit):
+def _update_commit(project, repo, commit):
     parser = os.path.join(_here, '../project/apps/patchwork/parser.py')
     cmd = 'git show %s | python %s -#' % (commit, parser)
-    hash = subprocess.check_output([cmd], cwd=gitdir, shell=True).strip()
+    hash = subprocess.check_output([cmd], cwd=repo.path, shell=True).strip()
 
     s = State.objects.get(name='Superseded')
     patches = Patch.objects.filter(project=project, hash=hash).exclude(state=s)
@@ -69,19 +36,11 @@ def _update_commit(project, gitdir, commit):
 def _update_project(repo_dir, project):
     log.info('Checking for updates to %s', project.linkname)
 
-    path = os.path.join(repo_dir, project.linkname)
-    if not os.path.exists(path):
-        _clone_repo(repo_dir, project.scm_url, project.linkname)
-    else:
-        _pull(path)
-
-    for commit in _commits_to_check(path):
-        _update_commit(project, path, commit)
+    repo = gitrepo.Repo(repo_dir, project.linkname, project.scm_url)
+    repo.update()
 
-    with open(os.path.join(path, '.git/refs/heads/master')) as f:
-        head = f.read()
-        with open(os.path.join(path, 'patchwork-last-commit'), 'w') as f:
-            f.write(head)
+    for commit in repo.process_unchecked_commits():
+        _update_commit(project, repo, commit)
 
 
 if __name__ == '__main__':