def test_new_parse_messages(self):
chash = '35d8e493ef66bd8c01c15a519c15d9a6d31cb2f4'
results = [i for i in parse_messages(self.repo_path, chash)]
assert len(results) == self.total_commits - 1
r = results[0]
assert r.hash_val == 'b6ca6229284e18b0ce8defeb4b240aa2f26223b4'
assert r.subject == 'added a function with friends.'
assert r.message_body == 'Co-authored-by: johnbaldwin <*****@*****.**>\nCo-authored-by: awong <[email protected]\n'
assert r.timestamp == '2019-02-26T18:22:54-05:00'
def repo_to_objects(owner: str, project: str, repopath: str, last_commit=""):
""" Retrieve objects from last commit if exists
This function is a generator so we can specify a buffer size
when making commits to the database. Otherwise, the I/O would
slow things way down.
:param repo_name: git user/git repo name, 'tbonza/EDS19'
:param dirpath: path to directory storing repo information
:param last_commit: string of git commit hash last stored in database
:return: generator of populated model database objects
:rtype: sqlalchemy database objects
"""
if len(last_commit) == 0:
cmts = parse_commits(repopath)
msgs = parse_messages(repopath)
fobjs = parse_committed_files(repopath)
else:
# retrieve from last commit HEAD
# need to set 'c1' lists
cmts = parse_commits(repopath, chash=last_commit)
msgs = parse_messages(repopath, chash=last_commit)
fobjs = parse_committed_files(repopath, chash=last_commit)
# map objects to database objects
m = re.compile(r"Co-authored-by\:(.*?)<(.*?)>")
for msg, cmt in zip(msgs, cmts):
contrib_id = 0
commit_authored_datetime = datetime.fromisoformat(
cmt.committer_timestamp)
yrmo_cad = str(commit_authored_datetime.year) + "-" + str(
commit_authored_datetime.month)
msg_item = Info(commit_hash=msg.hash_val,
subject=msg.subject,
message=msg.message_body,
created=datetime.fromisoformat(msg.timestamp))
meta_item = Meta(commit_hash=msg.hash_val,
owner_name=owner,
project_name=project,
yearmo=yrmo_cad)
author_item = Author(commit_hash=cmt.hash_val,
name=cmt.author,
email=cmt.author_email,
authored=commit_authored_datetime)
contrib_item = Contrib(contrib_id=contrib_id,
commit_hash=cmt.hash_val,
name=cmt.committer,
email=cmt.committer_email,
contributed=datetime.fromisoformat(
cmt.committer_timestamp))
yield msg_item
yield meta_item
yield author_item
yield contrib_item
# some commits will have multiple contributors
if m.match(msg.message_body):
contrib_id += 1
for item in m.findall(msg.message_body):
contrib_item = Contrib(contrib_id=contrib_id,
commit_hash=msg.hash_val,
name=item[0].strip(),
email=item[1].strip(),
contributed=datetime.fromisoformat(
msg.timestamp))
yield contrib_item
contrib_id += 1
file_id = 0
for fobj in fobjs:
cf_item = CommitFile(file_id=file_id,
commit_hash=fobj.hash_val,
modified_file=fobj.file_path,
lines_added=make_digit(fobj.added, "fobj.added"),
lines_subtracted=make_digit(
fobj.deleted, "fobj.deleted"))
yield cf_item
file_id += 1