def commit_tree_to_frame(commit):
tree_dicts = []
_add_subtree(tree_dicts, commit.tree, '.')
tree_frame = pd.DataFrame(tree_dicts)
tree_frame['hexsha'] = commit.hexsha
tree_frame['child_type'] = su.categorize(tree_frame['child_type'])
return tree_frame
def commits_to_frame(commits):
commit_attrs = (
'hexsha', 'name_rev', 'size',
'author.name', 'author.email',
'authored_datetime', 'author_tz_offset',
'committer.name', 'committer.email',
'committed_datetime', 'committer_tz_offset',
'encoding', 'message',
'stats.total.files', 'stats.total.lines',
'stats.total.insertions', 'stats.total.deletions',
'stats.files')
column_names = {attr: attr.replace('.', '_') for attr in commit_attrs}
commit_frame = su.dicts_to_dataframe(list(
su.objs_to_dicts(commits, commit_attrs)))
commit_frame.rename(columns=column_names, inplace=True)
commit_frame['name_rev'] = commit_frame['name_rev'].str.split(
' ', 1).apply(lambda x: x[-1])
categorical_cols = (
'name_rev', 'author_name', 'author_email',
'committer_name', 'committer_email', 'encoding')
for c in categorical_cols:
commit_frame[c] = su.categorize(commit_frame[c])
for c in ('authored_datetime', 'committed_datetime'):
commit_frame[c] = commit_frame[c].astype('datetime64[ns]')
commit_frame['message'] = commit_frame['message'].str.replace('\n', '\\n')
commit_frame = commit_frame.sort_values(
'committed_datetime', ascending=False).reset_index(drop=True)
return commit_frame
def commit_trees_to_frame(commits):
frame: pd.DataFrame = pd.concat(
(commit_tree_to_frame(c) for c in commits))
cat_columns = ('hexsha', 'tree', 'child', 'child_type')
for col in cat_columns:
frame[col] = su.categorize(frame[col])
frame.reset_index(inplace=True, drop=True)
return frame
def commits_to_actor_frame(commits):
attrs = ('name', 'email')
authors = extract_actors(commits, 'author', attrs)
committers = extract_actors(commits, 'committer', attrs)
actors = pd.merge(authors, committers, on=attrs, how='outer')
actors = actors.drop_duplicates().reset_index(drop=True).fillna(0)
for attr in attrs:
actors[attr] = su.categorize(actors[attr])
for col_name in ('author_commits', 'committer_commits'):
actors[col_name] = actors[col_name].astype('int')
return actors
def connect_actors(actor_frame, connectivity_sets, connectivity_column):
"""
:param actor_frame:
:param connectivity_sets:
:param connectivity_column:
:return:
Examples:
same_actors = {
'ccason': [3, 14, 15], 'clipka': [4, 5, 13],
'wfpokorny': [11, 17], 'anshuarya': [0],
'bentsm': [1], 'cbarton': [2], 'dbodor': [6],
'jlecher': [7], 'jgrimbert': [8], 'nalvarez': [9],
'selvik': [10], 'wverhelst': [12], 'gryken': [16],
'github': [18]}
actor_frame = connect_actors(actor_frame, same_actors, 'actor_id')
"""
connectivity = {}
for actor_id, connectivity_set in connectivity_sets.items():
for actor in connectivity_set:
connectivity[actor] = actor_id
actor_frame[connectivity_column] = su.categorize(pd.Series(connectivity))
return actor_frame