from config import *
import pandas as pd
import pickle
import xlearn as xl
from userProfile import *
import time
from utils import *


# 将device_id、city_id拼接到对应的城市热门日记表。注意：下面预测集特征顺序要与训练集保持一致
def feature_en(user_profile):
    file_name = DIRECTORY_PATH + "diaryTestSet/{0}DiaryTop3000.csv".format(user_profile['city_id'])
    data = pd.read_csv(file_name)

    data["device_id"] = user_profile['device_id']

    now = datetime.now()
    data["hour"] = now.hour
    data["minute"] = now.minute

    data.loc[data["hour"] == 0, ["hour"]] = 24
    data.loc[data["minute"] == 0, ["minute"]] = 60
    data["hour"] = data["hour"].astype("category")
    data["minute"] = data["minute"].astype("category")
    # 虽然预测y，但ffm转化需要y，并不影响预测结果
    data["y"] = 0
    data = data.drop("city_id", axis=1)
    print(data.head(1))
    print("特征工程处理结束")
    return data


# 把ffm.pkl load进来，将上面的表转化为ffm格式
def transform_ffm_format(df, device_id):
    file_path = DIRECTORY_PATH + "ffm_{0}_{1}.pkl".format(DATA_START_DATE, DATA_END_DATE)
    with open(file_path, "rb") as f:
        ffm_format_pandas = pickle.load(f)
        data = ffm_format_pandas.transform(df)
        now = datetime.now().strftime("%Y-%m-%d-%H-%M")
        print("ffm格式转化结束")
        predict_file_name = DIRECTORY_PATH + "result/{0}_{1}DiaryTop3000.csv".format(device_id, now)
        data.to_csv(predict_file_name, index=False,header=None)
        print("ffm写到服务器")
        return predict_file_name


# 将模型加载，预测，把预测日记的概率值按照降序排序，存到一个表里
def predict(user_profile):
    instance = feature_en(user_profile)
    instance_file_path = transform_ffm_format(instance, user_profile["device_id"])

    ffm_model = xl.create_ffm()
    ffm_model.setTest(instance_file_path)
    ffm_model.setSigmoid()
    ffm_model.predict(DIRECTORY_PATH + "model_{0}-{1}_lr{2}_lambda{3}.out".format(DATA_START_DATE,
                                                                                  DATA_END_DATE, lr, l2_lambda),
                      DIRECTORY_PATH + "result/{0}_output.txt".format(user_profile['device_id']))
    print("预测结束")
    predict_save_to_local(user_profile, instance)
    #TODO 没有提供生产环境的redis地址，所以这个函数先不运行
    # predict_save_to_redis(user_profile, instance)

# 将预测结果与device_id 进行拼接，并按照概率降序排序
def wrapper_result(user_profile, instance):
    proba = pd.read_csv(DIRECTORY_PATH +
                                "result/{0}_output.txt".format(user_profile['device_id']), header=None)
    proba = proba.rename(columns={0: "prob"})
    proba["cid"] = instance['cid']
    proba = proba.sort_values(by="prob", ascending=False)
    proba = proba.head(50)
    return proba

# 预测候选集保存到本地
def predict_save_to_local(user_profile, instance):
    proba = wrapper_result(user_profile, instance)
    proba.loc[:, "url"] = proba["cid"].apply(lambda x: "http://m.igengmei.com/diary_book/" + str(x[6:]) + '/')
    proba.to_csv(DIRECTORY_PATH + "result/feed_{}".format(user_profile['device_id']), index=False)
    print("成功将预测候选集保存到本地")

# 预测候选集保存到redis
def predict_save_to_redis(user_profile, instance):
    device_id = user_profile['device_id']
    cid_list = wrapper_result(user_profile, instance)["cid"].values.tolist()
    add_data_to_redis(device_id,cid_list)
    print("成功将预测候选集保存到redis")

def router(device_id):
    user_profile, not_exist = fetch_user_profile(device_id)

    if not_exist==1:
        print('Sorry, we don\'t have you.')
    else:
        predict(user_profile)


if __name__ == "__main__":
    # TODO 如果耗时小于一分钟，下一次取到的device_id和上一次相同
    while True:
        start = time.time()
        empty,device_id_list = get_active_users()
        if empty==1:
            time.sleep(60)
        else:
            old_device_id_list = pd.read_csv(DIRECTORY_PATH + "data_set_device_id.csv")["device_id"].values.tolist()
            for device_id in device_id_list:
                if device_id in old_device_id_list:
                    router(device_id)
                else:
                    print("该用户不是老用户，不能预测")
            end = time.time()
            time_cost = (end - start)
            print("耗时{}秒".format(time_cost))