def issues_network(out, repo, github):
"""Builds issues netowrk"""
interactions = datautil.get_issues_interaction(repo, github)
graph = networkutil.create_interaction_network(interactions,
repo_name=repo)
nx.write_adjlist(graph, out)
def commit_networks(in_pattern, out_dir, languages):
""" Builds commit network.
It retrieves commits from the top most-watched
repos of the given language. Then, for each repo,
it builds interaction network from the commit history
of each blob.
param:
in_pattern: The path pattern of file containing top
most watched repos.
out_dir: The pattern of the output dir for both
raw data and graph data.
lang: List of language.
Here is the example of each param:
in_pattern = "../data/most_watched/{0}.txt"
out_dir = "../data/network/commit/{0}"
lang = ['python', 'java']
It should be noted that in out_dir, it is expected
that there will be sub-dir graph/ and raw/.
"""
graph_out = out_dir + "/graph/{1}.txt"
raw_out = out_dir + "/raw/{1}.pickle"
github = Github(requests_per_second=1)
for lang in languages:
with open(in_pattern.format(lang), "r") as f:
for line in f:
repo = line.strip()
f_name = repo.replace('/', '_')
print "Starting {0} at {1}".format(repo, raw_out.format(
lang, f_name))
commits = get_commits_from_repo(repo, github)
with open(raw_out.format(lang, f_name), "w") as pickle_f:
pickle.dump(commits, pickle_f)
print "Starting {0} at {1}".format(repo, graph_out.format(
lang, f_name))
commit_interactions = commit_interactions_from_repo(commits)
g = networkutil.create_interaction_network(
commit_interactions, repo_name=repo)
nx.write_adjlist(g, graph_out.format(lang, f_name))