def setUp(self):
self.repo_path = mkdtemp()
rmdir(self.repo_path)
self.repo = get_repo(self.repo_path, VCS, create=True)
self.commit(self.commits)
self.project = Project.objects.create(slug='p1',
vcs_type=VCS,
repo_uri=self.repo_path)
def setUp(self):
self.repo_path = mkdtemp()
rmdir(self.repo_path)
self.repo = get_repo(self.repo_path, VCS, create=True)
self.commit(self.commits)
self.project = Project.objects.create(
slug='p1', vcs_type=VCS,
repo_uri=self.repo_path
)
def __get_codes_stats(repo_name):
repos_path = get_repos_path()
repo = get_repo(safe_str(os.path.join(repos_path, repo_name)))
tip = repo.get_changeset()
code_stats = {}
def aggregate(cs):
for f in cs[2]:
ext = lower(f.extension)
if ext in LANGUAGES_EXTENSIONS_MAP.keys() and not f.is_binary:
if ext in code_stats:
code_stats[ext] += 1
else:
code_stats[ext] = 1
map(aggregate, tip.walk('/'))
return code_stats or {}
def get_commits_stats(repo_name, ts_min_y, ts_max_y):
try:
log = get_commits_stats.get_logger()
except:
log = logging.getLogger(__name__)
lockkey = __get_lockkey('get_commits_stats', repo_name, ts_min_y,
ts_max_y)
lockkey_path = config['here']
log.info('running task with lockkey %s', lockkey)
try:
lock = l = DaemonLock(file_=jn(lockkey_path, lockkey))
#for js data compatibilty cleans the key for person from '
akc = lambda k: person(k).replace('"', "")
co_day_auth_aggr = {}
commits_by_day_aggregate = {}
repos_path = get_repos_path()
repo = get_repo(safe_str(os.path.join(repos_path, repo_name)))
repo_size = len(repo.revisions)
#return if repo have no revisions
if repo_size < 1:
lock.release()
return True
skip_date_limit = True
parse_limit = int(config['app_conf'].get('commit_parse_limit'))
last_rev = 0
last_cs = None
timegetter = itemgetter('time')
sa = get_session()
dbrepo = sa.query(Repository)\
.filter(Repository.repo_name == repo_name).scalar()
cur_stats = sa.query(Statistics)\
.filter(Statistics.repository == dbrepo).scalar()
if cur_stats is not None:
last_rev = cur_stats.stat_on_revision
if last_rev == repo.get_changeset().revision and repo_size > 1:
#pass silently without any work if we're not on first revision or
#current state of parsing revision(from db marker) is the
#last revision
lock.release()
return True
if cur_stats:
commits_by_day_aggregate = OrderedDict(json.loads(
cur_stats.commit_activity_combined))
co_day_auth_aggr = json.loads(cur_stats.commit_activity)
log.debug('starting parsing %s', parse_limit)
lmktime = mktime
last_rev = last_rev + 1 if last_rev > 0 else last_rev
for cs in repo[last_rev:last_rev + parse_limit]:
last_cs = cs # remember last parsed changeset
k = lmktime([cs.date.timetuple()[0], cs.date.timetuple()[1],
cs.date.timetuple()[2], 0, 0, 0, 0, 0, 0])
if akc(cs.author) in co_day_auth_aggr:
try:
l = [timegetter(x) for x in
co_day_auth_aggr[akc(cs.author)]['data']]
time_pos = l.index(k)
except ValueError:
time_pos = False
if time_pos >= 0 and time_pos is not False:
datadict = \
co_day_auth_aggr[akc(cs.author)]['data'][time_pos]
datadict["commits"] += 1
datadict["added"] += len(cs.added)
datadict["changed"] += len(cs.changed)
datadict["removed"] += len(cs.removed)
else:
if k >= ts_min_y and k <= ts_max_y or skip_date_limit:
datadict = {"time": k,
"commits": 1,
"added": len(cs.added),
"changed": len(cs.changed),
"removed": len(cs.removed),
}
co_day_auth_aggr[akc(cs.author)]['data']\
.append(datadict)
else:
if k >= ts_min_y and k <= ts_max_y or skip_date_limit:
co_day_auth_aggr[akc(cs.author)] = {
"label": akc(cs.author),
"data": [{"time":k,
"commits":1,
"added":len(cs.added),
"changed":len(cs.changed),
"removed":len(cs.removed),
}],
"schema": ["commits"],
}
#gather all data by day
if k in commits_by_day_aggregate:
commits_by_day_aggregate[k] += 1
else:
commits_by_day_aggregate[k] = 1
overview_data = sorted(commits_by_day_aggregate.items(),
key=itemgetter(0))
if not co_day_auth_aggr:
co_day_auth_aggr[akc(repo.contact)] = {
"label": akc(repo.contact),
"data": [0, 1],
"schema": ["commits"],
}
stats = cur_stats if cur_stats else Statistics()
stats.commit_activity = json.dumps(co_day_auth_aggr)
stats.commit_activity_combined = json.dumps(overview_data)
log.debug('last revison %s', last_rev)
leftovers = len(repo.revisions[last_rev:])
log.debug('revisions to parse %s', leftovers)
if last_rev == 0 or leftovers < parse_limit:
log.debug('getting code trending stats')
stats.languages = json.dumps(__get_codes_stats(repo_name))
try:
stats.repository = dbrepo
stats.stat_on_revision = last_cs.revision if last_cs else 0
sa.add(stats)
sa.commit()
except:
log.error(traceback.format_exc())
sa.rollback()
lock.release()
return False
#final release
lock.release()
#execute another task if celery is enabled
if len(repo.revisions) > 1 and CELERY_ON:
run_task(get_commits_stats, repo_name, ts_min_y, ts_max_y)
return True
except LockHeld:
log.info('LockHeld')
return 'Task with key %s already running' % lockkey
