def authenticate(self):
"""
authenticate()
Authenticates this application to github using
the cas-user git user credentials. This is hopefully temporary!
"""
s = requests.Session()
username = config["github"]["user"]
password = config["github"]["pass"]
s.auth = (username, password)
payload = {"scopes": ["repo"]}
r = s.get(self.request_auth, params=payload)
if r.headers.get('x-ratelimit-remaining') == '0':
logging.info("Github quota limit hit -- waiting")
# Wait up to a hour until we can continue..
while r.headers.get('x-ratelimit-remaining') == '0':
time.sleep(600) # Wait 10 minutes and try again
r = s.get(self.request_auth, params=payload)
data = r.json()
data = r.json()[0]
if r.status_code >= 400:
msg = data.get('message')
logging.error("Failed to authenticate issue tracker: \n" +msg)
return # Exit
else:
self.auth_token = data.get("token")
requests_left = r.headers.get('x-ratelimit-remaining')
logging.info("Analyzer has " + requests_left + " issue tracker calls left this hour")
def authenticate(self):
"""
authenticate()
Authenticates this application to github using
the cas-user git user credentials. This is hopefully temporary!
"""
s = requests.Session()
username = config["github"]["user"]
password = config["github"]["pass"]
s.auth = (username, password)
payload = {"scopes": ["repo"]}
r = s.get(self.request_auth, params=payload)
if r.headers.get('x-ratelimit-remaining') == '0':
logging.info("Github quota limit hit -- waiting")
# Wait up to a hour until we can continue..
while r.headers.get('x-ratelimit-remaining') == '0':
time.sleep(600) # Wait 10 minutes and try again
r = s.get(self.request_auth, params=payload)
data = r.json()
data = r.json()[0]
if r.status_code >= 400:
msg = data.get('message')
logging.error("Failed to authenticate issue tracker: \n" +msg)
return # Exit
else:
self.auth_token = data.get("token")
requests_left = r.headers.get('x-ratelimit-remaining')
logging.info("Analyzer has " + requests_left + " issue tracker calls left this hour")
def notify(self):
"""
Notify all subscribers that repo has been analyzed and is ready
to be viewed
"""
FROM = "*****@*****.**"
TO = self.subscribers
SUBJECT = "Your repository has been analyzed"
TEXT = "Your analyzed repository is now ready to be viewed at http://kiwi.se.rit.edu/repo/" + self.repo
# prepare actual message
message = """\From: %s\nTo: %s\nSubject: %s\n\n%s""" % (
FROM, ", ".join(TO), SUBJECT, TEXT)
try:
server = smtplib.SMTP("smtp.gmail.com", 587)
server.ehlo()
server.starttls()
server.login(self.gmail_user, self.gmail_pwd)
server.sendmail(FROM, TO, message)
server.quit()
logging.info("Notification sent successfully")
except:
logging.error("Failed to send notification")
def notify(self):
"""
Notify all subscribers that repo has been analyzed and is ready
to be viewed
"""
FROM = "*****@*****.**"
TO = self.subscribers
SUBJECT = "Your repository has been analyzed"
TEXT = "Your analyzed repository is now ready to be viewed at http://kiwi.se.rit.edu/repo/" + self.repo
# prepare actual message
message = """\From: %s\nTo: %s\nSubject: %s\n\n%s""" % (FROM, ", ".join(TO), SUBJECT, TEXT)
try:
server = smtplib.SMTP("smtp.gmail.com", 587)
server.ehlo()
server.starttls()
server.login(self.gmail_user, self.gmail_pwd)
server.sendmail(FROM, TO, message)
server.quit()
logging.info("Notification sent successfully")
except:
logging.error("Failed to send notification")
def getDateOpened(self, issueNumber):
"""
getDateOpened()
Gets the date the issue number was opened in unix time
If issue cannot be found for whichever reason, returns null.
"""
logging.info("searching for issue: " + str(issueNumber))
logging.info(self.request_repos + "/" + self.owner + "/" + self.repo +
"/issues/" + issueNumber)
header = {'Authorization': 'token ' + self.auth_token}
r = requests.get(self.request_repos + "/" + self.owner + "/" +
self.repo + "/issues/" + issueNumber,
headers=header)
data = r.json()
# If forbidden
if r.status_code == 403:
# Check the api quota
if r.headers.get('x-ratelimit-remaining') == '0':
logging.info("Github quota limit hit -- waiting")
# Wait up to a hour until we can continue..
while r.headers.get('x-ratelimit-remaining') == '0':
time.sleep(600) # Wait 10 minutes and try again
r = requests.get(self.request_repos + "/" + self.owner +
"/" + self.repo + "/issues/" +
issueNumber,
headers=header)
data = r.json()
# Check for other error codes
elif r.status_code >= 400:
msg = data.get('message')
logging.info("issue not found")
return None
else:
try:
date = (dateutil.parser.parse(
data.get('created_at'))).timestamp()
return date
except:
logging.error(
"ISSUE TRACKER FAILURE: Could not get created_at from github issues API"
)
return None
def getDateOpened(self, issueNumber):
"""
getDateOpened()
Gets the date the issue number was opened in unix time
If issue cannot be found for whichever reason, returns null.
"""
logging.info("searching for issue: " + str(issueNumber))
logging.info(self.request_repos + "/" + self.owner + "/" +
self.repo + "/issues/" + issueNumber)
header = {'Authorization': 'token ' + self.auth_token}
r = requests.get(self.request_repos + "/" + self.owner + "/" +
self.repo + "/issues/" + issueNumber, headers=header)
data = r.json()
# If forbidden
if r.status_code == 403:
# Check the api quota
if r.headers.get('x-ratelimit-remaining') == '0':
logging.info("Github quota limit hit -- waiting")
# Wait up to a hour until we can continue..
while r.headers.get('x-ratelimit-remaining') == '0':
time.sleep(600) # Wait 10 minutes and try again
r = requests.get(self.request_repos + "/" + self.owner + "/" +
self.repo + "/issues/" + issueNumber, headers=header)
data = r.json()
# Check for other error codes
elif r.status_code >= 400:
msg = data.get('message')
logging.info("issue not found")
return None
else:
try:
date = (dateutil.parser.parse(data.get('created_at'))).timestamp()
return date
except:
logging.error("ISSUE TRACKER FAILURE: Could not get created_at from github issues API")
return None
def calculateCommitRiskyness(self, commits, coefficient_names):
"""
calcualte the probability of commits to be buggy or not
using the linear regression model
estimated probability = 1/[1 + exp(-a - BX)]
"""
# 2 cases: model cannot possibly be build if no signficant coefficients available
# in this case, we just insert -1 for the probability to indicate no glm prediction possible
if len(coefficient_names) == 0:
coefficient_dict = {}
model_available = False
else:
coefficient_dict = self._getCoefficients(coefficient_names)
model_available = True
intercept_value, intercept_pvalue = self._getInterceptValue(
coefficient_names)
for commit in commits:
if model_available == False:
commit.glm_probability = -1
else:
coefs_sum = 0
for coef_name, coef_value in coefficient_dict.items():
coefs_sum += (coef_value * getattr(commit, coef_name))
try:
riskyness = 1 / (1 +
math.exp(-intercept_value - coefs_sum))
except OverflowError:
logging.error("Overflow error for repo " + self.repo_id)
logging.error("Calculating riskyness for " +
commit.commit_hash)
logging.error("Sum of coefficients: " + str(coefs_sum))
logging.error("Coeffiecents: " + str(coefficient_dict))
riskyness = 0.01
commit.glm_probability = riskyness
def calculateCommitRiskyness(self, commits, coefficient_names):
"""
calcualte the probability of commits to be buggy or not
using the linear regression model
estimated probability = 1/[1 + exp(-a - BX)]
"""
# 2 cases: model cannot possibly be build if no signficant coefficients available
# in this case, we just insert -1 for the probability to indicate no glm prediction possible
if len(coefficient_names) == 0:
coefficient_dict = {}
model_available = False
else:
coefficient_dict = self._getCoefficients(coefficient_names)
model_available = True
intercept_value, intercept_pvalue = self._getInterceptValue(coefficient_names)
for commit in commits:
if model_available == False:
commit.glm_probability = -1
else:
coefs_sum = 0
for coef_name, coef_value in coefficient_dict.items():
coefs_sum += (coef_value * getattr(commit, coef_name))
try:
riskyness = 1/(1+ math.exp(-intercept_value-coefs_sum))
except OverflowError:
logging.error("Overflow error for repo " + self.repo_id)
logging.error("Calculating riskyness for " + commit.commit_hash)
logging.error("Sum of coefficients: " + str(coefs_sum))
logging.error("Coeffiecents: " + str(coefficient_dict))
riskyness = 0.01
commit.glm_probability = riskyness
for repo in testRepos:
session.merge(Repository(repo))
session.commit()
logging.info("Done.")
elif arg == "":
# No args, just do scan
logging.info("Starting Scan...")
repoSession = Session()
# Latest time to get new repo data (1 day ago)
refresh_date = str(datetime.utcnow() - timedelta(days=1))
# Get un-injested repos or repos not been updated since the refresh_date
reposToGet = (
repoSession.query(Repository)
.filter((Repository.ingestion_date == None) | (Repository.ingestion_date < refresh_date))
.all()
)
# TODO: This: (downloading and parsing commit logs
if len(reposToGet) > 0:
for repo in reposToGet:
localRepo = LocalRepository(repo)
localRepo.sync()
repoSession.merge(repo)
repoSession.commit()
logging.info("Done, finished everything.")
else:
logging.info("Nothing to do. Done.")
else:
logging.error("Invalid Command")
]
session = Session()
for repo in testRepos:
session.merge(Repository(repo))
session.commit()
logging.info('Done.')
elif arg == '':
# No args, just do scan
logging.info('Starting Scan...')
repoSession = Session()
# Latest time to get new repo data (1 day ago)
refresh_date = str(datetime.utcnow() - timedelta(days=1))
# Get un-injested repos or repos not been updated since the refresh_date
reposToGet = (repoSession.query(
Repository).filter((Repository.ingestion_date == None)
| (Repository.ingestion_date < refresh_date)).all())
#TODO: This: (downloading and parsing commit logs
if len(reposToGet) > 0:
for repo in reposToGet:
localRepo = LocalRepository(repo)
localRepo.sync()
repoSession.merge(repo)
repoSession.commit()
logging.info('Done, finished everything.')
else:
logging.info('Nothing to do. Done.')
else:
logging.error('Invalid Command')
