def parseEmail(emailFile):
escon = es.ESload()
start = time.time()
f = open(emailFile, 'r')
msg = email.message_from_file(f)
f.close()
env = open(emailFile + '.envelope', 'r')
envelope = env.readline()
env.close()
metaData = getMetaData(msg)
try:
if msg.is_multipart():
data = mimeEmail(msg, envelope, metaData, )
else:
data = rawEmail(msg, envelope, metaData)
except NoBoundaryInMultipartDefect:
data = rawEmail(msg, envelope, metaData, )
key = createKey(metaData[0], envelope)
envData = elasticEnvelope(envelope)
#write data into ES
escon.indexEmailData(data, key)
escon.indexEnvelopeData(envData, key)
duration = time.time() - start
return duration
def checkData(emailFile):
logger = insertData.get_logger()
logger.info("Reading email %s" % emailFile)
logger.info("Task id %s" % checkData.request.id)
env = open(emailFile + '.envelope', 'r')
envelope = env.readline()
env.close()
f = open(emailFile, 'r')
msg = message_from_file(f)
metaData = getMetaData(msg)
key = createKey(metaData[0], envelope)
f.close()
email = rawEmail(key)
#print email
print key
if email != 0:
f = open(emailFile, 'r')
sEmail = f.readlines()
f.close()
m = hashlib.sha1()
m.update(''.join(sEmail))
sHash = m.hexdigest()
m = hashlib.sha1()
m.update(email)
dHash = m.hexdigest()
if sHash != dHash:
logger.info("[Error/EmailTest] Email corrupted.] < %s >" % emailFile)
#return "0"
else:
logger.info("[Email data test: OK]")