"""
股性分析
"""

# 是否有涨停
import copy
import json
import random
import time

from code_attribute import gpcode_manager

# 代码股性记录管理
from db import redis_manager
from utils import tool
from db.redis_manager_delegate import RedisManager, RedisUtils
from utils.tool import CodeDataCacheUtil


class CodeNatureRecordManager:
    __k_format_cache = {}
    __nature_cache = {}

    __db = 0
    __instance = None
    __redis_manager = redis_manager.RedisManager(0)

    def __new__(cls, *args, **kwargs):
        if not cls.__instance:
            cls.__instance = super(CodeNatureRecordManager, cls).__new__(cls, *args, **kwargs)
            cls.__load_datas()
        return cls.__instance

    @classmethod
    def __get_redis(cls):
        return cls.__redis_manager.getRedis()

    @classmethod
    def __load_datas(cls):
        __redis = cls.__get_redis()
        try:
            keys = RedisUtils.keys(__redis, "k_format-*")
            for k in keys:
                code = k.split("-")[1]
                val = RedisUtils.get(__redis, k)
                val = json.loads(val)
                cls.__k_format_cache[code] = val
            keys = RedisUtils.keys(__redis, "code_nature-*")
            for k in keys:
                code = k.split("-")[1]
                val = RedisUtils.get(__redis, k)
                val = json.loads(val)
                cls.__nature_cache[code] = val
        except:
            pass
        finally:
            RedisUtils.realse(__redis)

    # 保存K线形态
    def save_k_format(self, code, k_format):
        self.__k_format_cache[code] = k_format
        RedisUtils.setex_async(self.__db, f"k_format-{code}", tool.get_expire(), json.dumps(k_format))

    def get_k_format(self, code):
        val = RedisUtils.get(self.__get_redis(), f"k_format-{code}")
        if val:
            return json.loads(val)
        return None

    def get_k_format_cache(self, code):
        val = self.__k_format_cache.get(code)
        # 复制
        return copy.deepcopy(val) if val else None

    def clear(self):
        self.__k_format_cache.clear()
        self.__nature_cache.clear()
        keys = RedisUtils.keys(self.__get_redis(), "k_format-*")
        for k in keys:
            RedisUtils.delete(self.__get_redis(), k)
        keys = RedisUtils.keys(self.__get_redis(), "code_nature-*")
        for k in keys:
            RedisUtils.delete(self.__get_redis(), k)

    # 保存股性

    def save_nature(self, code, natures):
        RedisUtils.setex_async(self.__db, f"code_nature-{code}", tool.get_expire(), json.dumps(natures))

    def get_nature(self, code):
        val = RedisUtils.get(self.__get_redis(), f"code_nature-{code}")
        if val:
            return json.loads(val)
        return None

    def get_nature_cache(self, code):
        return self.__nature_cache.get(code)


class LatestMaxVolumeManager:
    __db = 0
    __instance = None
    __redis_manager = redis_manager.RedisManager(0)
    __max_volume_cache = {}

    def __new__(cls, *args, **kwargs):
        if not cls.__instance:
            cls.__instance = super(LatestMaxVolumeManager, cls).__new__(cls, *args, **kwargs)
            cls.__load_datas()
        return cls.__instance

    @classmethod
    def __load_datas(cls):
        __redis = cls.__get_redis()
        try:
            keys = RedisUtils.keys(__redis, "is_latest_max_volume-*")
            for k in keys:
                code = k.split("-")[-1]
                val = RedisUtils.get(__redis, k)
                CodeDataCacheUtil.set_cache(cls.__max_volume_cache, code, val)
        finally:
            RedisUtils.realse(__redis)

    @classmethod
    def __get_redis(cls):
        return cls.__redis_manager.getRedis()

    def __save_has_latest_max_volume(self, code):
        RedisUtils.setex_async(self.__db, f"is_latest_max_volume-{code}", tool.get_expire(), "1")

    # 设置最近有最大量
    def set_has_latest_max_volume(self, code):
        CodeDataCacheUtil.set_cache(self.__max_volume_cache, code, 1)
        self.__save_has_latest_max_volume(code)

    # 最近是否有最大量
    def is_latest_max_volume(self, code):
        return code in self.__max_volume_cache

    def clear(self):
        self.__max_volume_cache.clear()
        keys = RedisUtils.keys(self.__get_redis(), "is_latest_max_volume-*")
        for k in keys:
            RedisUtils.delete_async(self.__db, k)


# 涨幅过高的票管理
class HighIncreaseCodeManager:
    __db = 0
    __instance = None
    __redis_manager = redis_manager.RedisManager(0)
    __high_increase_codes = set()

    def __new__(cls, *args, **kwargs):
        if not cls.__instance:
            cls.__instance = super(HighIncreaseCodeManager, cls).__new__(cls, *args, **kwargs)
            cls.__load_datas()
        return cls.__instance

    @classmethod
    def __load_datas(cls):
        __redis = cls.__get_redis()
        try:
            cls.__high_increase_codes = RedisUtils.smembers(__redis, "high_increase_codes")
        finally:
            RedisUtils.realse(__redis)

    @classmethod
    def __get_redis(cls):
        return cls.__redis_manager.getRedis()

    def add_code(self, code):
        if self.__high_increase_codes is None:
            self.__high_increase_codes = set()
        self.__high_increase_codes.add(code)
        RedisUtils.sadd_async(self.__db, "high_increase_codes", code)
        RedisUtils.expire_async(self.__db, "high_increase_codes", tool.get_expire())

    def is_in(self, code):
        return code in self.__high_increase_codes

    # 加载全部
    def list_all(self):
        return self.__high_increase_codes

    def clear(self):
        if self.__high_increase_codes:
            self.__high_increase_codes.clear()
        RedisUtils.delete_async(self.__db, "high_increase_codes")


# 设置历史K线
def set_record_datas(code, limit_up_price, record_datas):
    k_format = get_k_format(float(limit_up_price), record_datas)
    CodeNatureRecordManager().save_k_format(code, k_format)
    natures = get_nature(record_datas)
    CodeNatureRecordManager().save_nature(code, natures)


# 获取K线形态
# 返回 (15个交易日涨幅是否大于24.9%,是否破前高，是否超跌，是否接近前高，是否N,是否V,是否有形态,天量大阳信息,是否具有辨识度)
def get_k_format(limit_up_price, record_datas):
    p1_data = get_lowest_price_rate(record_datas)
    p1 = p1_data[0] >= 0.249, p1_data[1]
    p2 = __is_new_top(limit_up_price, record_datas)
    p3 = __is_lowest(record_datas)
    p4 = __is_near_new_top(limit_up_price, record_datas)
    p5 = __is_n_model(record_datas)
    p6 = __is_v_model(record_datas)
    p8 = __get_big_volumn_info(record_datas)

    # # N字型包含了N字型
    # if p5:
    #     p6 = False, ''

    p7 = (p1[0] or p2[0] or p3[0] or p4[0] or p5[0] or p6[0], '')

    # 是否具有辨识度
    p9 = is_special(record_datas)
    p10 = is_latest_10d_max_volume_at_latest_2d(record_datas)

    return p1, p2, p3, p4, p5, p6, p7, p8, p9, p10


# 是否具有K线形态
def is_has_k_format(limit_up_price, record_datas):
    is_too_high, is_new_top, is_lowest, is_near_new_top, is_n, is_v, has_format, volume_info, is_special = get_k_format(
        float(limit_up_price), record_datas)
    if not has_format:
        return False, "不满足K线形态"
    return True, "有形态"


# 获取股性
# 返回（是否涨停，首板溢价率，首板炸板溢价率）
def get_nature(record_datas):
    limit_up_count = get_first_limit_up_count(record_datas)
    premium_rate = get_limit_up_premium_rate(record_datas)
    open_premium_rate = get_open_limit_up_premium_rate(record_datas)
    result = (limit_up_count, premium_rate, open_premium_rate)
    return result


# 获取涨幅
def get_lowest_price_rate(record_datas):
    datas = copy.deepcopy(record_datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-10:]
    for data in datas:
        limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(data["pre_close"]))
        if abs(limit_up_price - data["high"]) < 0.01:
            date = data['bob'].strftime("%Y-%m-%d")
            return round((datas[-1]["close"] - data["close"]) / data["close"], 4), date
    return 0, ''


# 是否涨得太高
def is_up_too_high_in_10d(record_datas):
    datas = copy.deepcopy(record_datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-10:]
    limit_ups = []
    limit_up_count = 0
    max_price = datas[0]["high"]
    for data in datas:
        limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(data["pre_close"]))
        date = data['bob'].strftime("%Y-%m-%d")
        if data["high"] > max_price:
            max_price = data["high"]
        if abs(limit_up_price - data["high"]) < 0.01:
            limit_ups.append((date, True))
            limit_up_count += 1
        else:
            limit_ups.append((date, False))
    # 下降幅度
    limit_down_rate = round((max_price - datas[-1]["close"]) / max_price, 3)

    if limit_up_count < 3:
        return False
    # 连续5天有3天涨停
    for i in range(len(limit_ups)):
        if i + 5 > len(limit_ups):
            break
        temp_datas = limit_ups[i:i + 5]
        t_count = 0
        for t in temp_datas:
            if t[1]:
                t_count += 1
        if t_count >= 3 and limit_down_rate < 0.15:
            # 降幅小于20%
            return True

    return False


# 10天内的最高量是否集中在最近两天
def is_latest_10d_max_volume_at_latest_2d(record_datas):
    datas = copy.deepcopy(record_datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-10:]
    max_volume_info = None
    for i in range(0, len(datas)):
        if not max_volume_info:
            max_volume_info = (i, datas[i]["volume"])
        else:
            if max_volume_info[1] < datas[i]["volume"]:
                max_volume_info = (i, datas[i]["volume"])
    return len(datas) - max_volume_info[0] <= 2


# 120 天内是否长得太高
def is_up_too_high_in_120d(record_datas):
    datas = copy.deepcopy(record_datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-120:]
    today_limit_up_price = round(float(gpcode_manager.get_limit_up_price_by_preprice(datas[-1]["close"])), 2)
    max_price = 0
    for data in datas:
        if data["high"] > max_price:
            max_price = data["high"]
    if today_limit_up_price <= max_price:
        return False

    # 计算120天的均价
    total_amount = 0
    total_volume = 0
    for data in datas:
        total_amount += data["amount"]
        total_volume += data["volume"]
    average_price = round(total_amount / total_volume, 2)
    if (today_limit_up_price - average_price) / average_price > 0.3:
        return True
    else:
        return False


# 最近几天是否有最大量
def is_have_latest_max_volume(record_datas, day_count):
    datas = copy.deepcopy(record_datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-120:]
    max_volume = (0, datas[0]["volume"])
    for i in range(0, len(datas)):
        if max_volume[1] < datas[i]["volume"]:
            max_volume = (i, datas[i]["volume"])
    if len(datas) - max_volume[0] <= day_count:
        return True
    return False


# 是否有涨停
def get_first_limit_up_count(datas):
    datas = copy.deepcopy(datas)
    count = 0
    for i in range(len(datas)):
        item = datas[i]
        # 获取首板涨停次数
        if __is_limit_up(item) and i > 0 and not __is_limit_up(datas[i - 1]):
            # 首板涨停
            count += 1

    return count


# 是否破前高
def __is_new_top(limit_up_price, datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-80:]
    max_price = 0
    for data in datas:
        if max_price < data["high"]:
            max_price = data["high"]
    if limit_up_price >= max_price:
        return True, ''
    return False, ''


def is_new_top(limit_up_price, datas):
    return __is_new_top(float(limit_up_price), datas)[0]


# 接近新高
def __is_near_new_top(limit_up_price, datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-80:]

    max_price = 0
    max_price_index = 0
    for index in range(0, len(datas)):
        data = datas[index]
        if data["high"] > max_price:
            max_price = data["high"]
            max_price_index = index
    rate = (max_price - float(limit_up_price)) / max_price
    if 0 < rate < 0.02:
        return True, datas[max_price_index]['bob'].strftime("%Y-%m-%d")
    return False, ''


# 是否跌破箱体
def __is_lowest(datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-80:]
    min_price = 100000
    for data in datas:
        if min_price > data["low"]:
            min_price = data["low"]
    # 近5天内的最低价
    date = ''
    min_price_5 = 10000
    for data in datas[-5:]:
        if min_price_5 > data["low"]:
            min_price_5 = data["low"]
            date = data['bob']
    if abs(min_price_5 - min_price) / min_price < 0.015:
        return True, date.strftime("%Y-%m-%d")
    return False, ''


# N字形
def __is_n_model(datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-80:]
    if len(datas) >= 6:
        max_price = 0
        min_price = 1000000
        for i in range(len(datas) - 5, len(datas)):
            item = datas[i]
            print(item)
            limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(item["pre_close"]))
            if abs(limit_up_price - item["high"]) < 0.001 and abs(
                    limit_up_price - datas[i - 1]["high"]) >= 0.001:
                # 涨停，前一天非涨停
                max_price = item["close"]
            elif max_price > 0:
                if min_price > item["low"]:
                    min_price = item["low"]
        if max_price > min_price:
            return True, ''
    return False, ''


# V字形
def __is_v_model(datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-30:]
    max_price = 0
    max_price_index = -1
    for i in range(0, len(datas)):
        if max_price < datas[i]["close"]:
            max_price = datas[i]["close"]
            max_price_index = i
    min_price = max_price
    min_price_index = max_price_index
    for i in range(max_price_index, len(datas)):
        if min_price > datas[i]["close"]:
            min_price = datas[i]["close"]
            min_price_index = i

    if (max_price - min_price) / max_price > 0.249:
        return True, ''

    return False, ''


# 是否天量大阳
def __get_big_volumn_info(datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    datas = datas[-30:]
    max_volume = 0
    total_volume = 0
    for data in datas:
        if max_volume < data["volume"]:
            max_volume = data["volume"]
        total_volume += data["volume"]
    average_volume = total_volume // len(datas)
    return max_volume, average_volume


# 是否涨停
def __is_limit_up(data):
    limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(data["pre_close"]))
    return abs(limit_up_price - data["close"]) < 0.001


# 是否涨停过
def __is_limited_up(data):
    limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(data["pre_close"]))
    return abs(limit_up_price - data["high"]) < 0.001


# 首板涨停溢价率
def get_limit_up_premium_rate(datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    first_rate_list = []
    for i in range(0, len(datas)):
        item = datas[i]
        limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(item["pre_close"]))
        if abs(limit_up_price - item["close"]) < 0.001:
            if 0 < i < len(datas) - 1 and not __is_limit_up(datas[i - 1]):
                # 首板涨停
                rate = (datas[i + 1]["high"] - datas[i + 1]["pre_close"]) / datas[i + 1]["pre_close"]
                first_rate_list.append(rate)
    if not first_rate_list:
        return None
    count = 0
    for rate in first_rate_list:
        if rate >= 0.01:
            count += 1
    return count / len(first_rate_list)


# 首板炸板溢价率
def get_open_limit_up_premium_rate(datas):
    datas = copy.deepcopy(datas)
    datas.sort(key=lambda x: x["bob"])
    first_rate_list = []
    for i in range(0, len(datas)):
        item = datas[i]
        limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(item["pre_close"]))

        if abs(limit_up_price - item["high"]) < 0.001 and abs(limit_up_price - item["close"]) > 0.001:
            #
            limit_up_price = float(gpcode_manager.get_limit_up_price_by_preprice(datas[i - 1]["pre_close"]))
            if 0 < i < len(datas) - 1 and not __is_limit_up(datas[i - 1]):
                # 前一天未涨停
                rate = (datas[i + 1]["high"] - item["high"]) / item["high"]
                first_rate_list.append(rate)
    if not first_rate_list:
        return None
    count = 0
    for rate in first_rate_list:
        if rate >= 0.01:
            count += 1
    return count / len(first_rate_list)


# 是否具有辨识度
def is_special(datas):
    # 30个交易日内有≥5天曾涨停且连续涨停数或曾涨停≥2天
    if len(datas) > 30:
        datas = datas[-30:]

    count = 0
    continue_count = 0
    last_index = -1
    for i in range(len(datas)):
        if __is_limited_up(datas[i]):
            if last_index >= 0 and i - last_index == 1:
                continue_count += 1
            count += 1
            last_index = i
    if count >= 5 and continue_count > 0:
        return True, ''
    return False, ''


if __name__ == "__main__":
    HighIncreaseCodeManager().add_code("000333")
    print(HighIncreaseCodeManager().is_in("000333"))
    print(HighIncreaseCodeManager().is_in("000222"))