dbpedia_run_log_http.py 文件源码

def run_log(query_line, last_timestamp):
    # open queries and regex for links
    url_ = re.findall('"GET (.*?) HTTP', query_line)

    last_timestamp_new = datetime.utcnow()
    if len(url_) == 1:
        request_url = url_[0]
        query_times = []
        resp = ''
        result_size = 0

        try:
            utcnow = datetime.utcnow()
            midnight_utc = datetime.combine(utcnow.date(), time(0))

            delta_last_query = (datetime.utcnow() - last_timestamp).total_seconds()

            for _ in range(11):
                response, exec_time = run_http_request(request_url)
                # if exec_time == -1.:
                #   break

                query_times.append(exec_time)
                # timesleep.sleep(random.random()*0.1)

            last_timestamp_new = datetime.utcnow()
            timestamp_query = ((last_timestamp_new - midnight_utc).total_seconds())

            respJson = response.json()
            result_size = get_result_size(respJson)

        except:
            exec_time = -1

        if exec_time != -1 and len(query_times) == 11: #and result_size > 0:
            cold_exec_time = query_times[0]
            warm_times = query_times[1:]
            warm_mean = np.mean(warm_times, dtype=np.float64)

            time_vec = [timestamp_query, delta_last_query]

            query_clean = cleanup_query(request_url)

            res = str(query_clean + '\t'+ str(time_vec) + '\t' + str(warm_mean) + '\t' + str(cold_exec_time) + '\t' + str(result_size) + '\n')
            return (res, last_timestamp_new)

        else:
            return (-1., last_timestamp_new)
    else:
        return (-1., last_timestamp_new)