def standardize_repos_and_authors(self):
""""
Doesn't modify a scope, but returns some plural forms that let callers not know' \
if the scope specified plural or singular objects
"""
repos = None
authors = None
if self.author:
if isinstance(self.author, str):
email = self.author.replace(" ", "+")
authors = Author.objects.filter(email=email)
else:
authors = Author.objects.filter(pk=self.author.pk).values_list(
'pk', flat=True)
elif self.repo:
authors = self.repo.author_ids(self.start, self.end)
else:
authors = None
if not self.repo and self.author:
repos = Author.repos(repos, self.start,
self.end).values_list('pk', flat=True)
elif self.repo:
repos = Repository.objects.filter(pk=self.repo.pk).values_list(
'pk', flat=True)
else:
assert self.org is not None
repos = [
x.pk for x in Repository.objects.filter(
organization=self.org).all()
]
return (repos, authors)
def repo_table(scope): # repos, start, end):
#this drives the list of all repos within an organization, showing the statistics for them within the selected
#time range, along with navigation links.
(repos, authors) = scope.standardize_repos_and_authors()
interval = 'DY'
# FIXME: explain
repos = [x.pk for x in scope.available_repos.all()]
stats = Statistic.queryset_for_range(repos=repos,
authors=authors,
start=scope.start,
end=scope.end,
interval=interval)
stats = Statistic.annotate(
stats.values('repo__name')).order_by('repo__name')
data = _annotations_to_table(stats, 'repo', 'repo__name')
# FIXME: insert in author count, which is ... complicated ... this can be optimized later
# we should be able to grab every repo and annotate it with the author count in one extra query tops
# but it might require manually writing it.
for d in data:
repo = d['repo']
author_count = Author.author_count(repo,
start=scope.start,
end=scope.end)
d['author_count'] = author_count
# some repos won't have been scanned, and this requires a second query to fill them into the table
repos = Repository.objects.filter(last_scanned=None,
organization=scope.org)
for unscanned in repos:
data.append(dict(repo=unscanned.name))
return data
def author_stats_table(scope, limit=None):
"""
this drives the author tables, both ranged and non-ranged, accessed off the main repo list.
the interval 'LF' shows lifetime stats, but elsewhere we just do daily roundups, so this parameter
should really be a boolean. The limit parameter is not yet used.
"""
results = []
interval = scope.interval
if interval != 'LF':
interval = 'DY'
authors = None
if scope.repo:
authors = Author.authors(scope.repo, scope.start, scope.end)
def add_stat(author, repo):
stat1 = Statistic.queryset_for_range(repo,
author=author,
start=scope.start,
end=scope.end,
interval=scope.interval)
stat2 = Statistic.compute_interval_statistic(stat1,
interval=scope.interval,
repo=repo,
author=author,
start=scope.start,
end=scope.end)
stat2 = stat2.to_dict()
stat2['author'] = author.email
stat2['repo'] = repo.name
if stat2['lines_changed']:
# skip authors with no contribution in the time range
results.append(stat2)
if not scope.author and scope.repo:
for author in authors:
add_stat(author, scope.repo)
elif scope.author and not scope.repo:
for repo in scope.author.repos():
add_stat(scope.author, repo)
elif scope.author and scope.repo:
add_stat(scope.author, scope.repo)
return results
def compute_daily_rollup(cls, repo=None, author=None, start_day=None, total_instances=None):
"""
Generate rollup stats for everything the team did on a given day
"""
end_date = cls.get_end_day(start_day, DAY)
file_change_count = FileChange.change_count(repo, author=author, start=start_day, end=end_date)
if file_change_count == 0:
# this looks like a merge commit, FIXME: it would be a good idea to validate that this is 100% true.
# print("-- skipping potential merge commit --")
return
if not author:
authors_count = Author.author_count(repo, start=start_day, end=end_date)
else:
authors_count = 1
# Aggregate values from query set for rollup
data = FileChange.aggregate_stats(repo, author=author, start=start_day, end=end_date)
# FIXME: if start_day is today, we need to UPDATE the current stat? - verify if the bulk_update code deals with this?
# FIXME: model code method below is rather inefficient, does this matter?
# Create total rollup row for the day
stat = Statistic(
start_date=start_day,
interval=DAY,
repo=repo,
author=author,
lines_added=data['lines_added'],
lines_removed=data['lines_removed'],
lines_changed=data['lines_changed'],
commit_total= data['commit_total'],
files_changed=data['files_changed'],
author_total=authors_count,
days_active=1,
)
stat.compute_derived_values()
cls.smart_bulk_update(repo=repo, start_day=start_day, author=author, interval=DAY, stat=stat, total_instances=total_instances)
def get_authors_for_repo(cls, repo):
"""
Return the authors involved in the repo
"""
return Author.authors(repo)