#下載資料套件
import urllib3
from bs4 import BeautifulSoup
import requests as r

#資料處理套件
import json
import pandas as pd
from datetime import datetime, date, timedelta

# 畫圖套件
import plotly.graph_objects as go
from plotly.subplots import make_subplots

def date_range(start_year, start_month, start_day, end_year, end_month, end_day):
    # 將開始日跟結束日轉為時間型態
    start_date = date(start_year, start_month, start_day)
    end_date = date(end_year, end_month, end_day)
    
    # 在開始日與結束日區間內 以每30天為單位切分 不含結束日
    date_list = pd.date_range(start_date, end_date, freq='30D').tolist()
    
    # 加上結束日 
    if date_list[-1] != end_date:
        date_list.append(end_date)
    
    return date_list

def pc_ratio(date_list):
    # 設定urllib物件,下載網址,儲存資料的df
    http = urllib3.PoolManager()
    url = "https://www.taifex.com.tw/cht/3/pcRatio"
    df = pd.DataFrame()
    
    # 下載期交所P/C比的迴圈
    for idx in range(len(date_list) - 1):
        # 設定下載資料的起始日 queryStartDate & queryEndDate
        if idx == 0:
            start_date = date_list[idx].strftime("%Y/%m/%d")
        else:
            start_date = (date_list[idx] + timedelta(days = 1)).strftime("%Y/%m/%d")
        end_date = date_list[idx + 1].strftime("%Y/%m/%d")
        print("迴圈索引值: ", idx, "; 下載時間區間: ", start_date, "-",end_date)
        
        # 從期交所下載資料
        res = http.request(
             'POST',
              url,
              fields={
                 'queryStartDate': start_date,
                 'queryEndDate': end_date
              }
         )
        html_doc = res.data
        table = BeautifulSoup(html_doc, 'html.parser').table
        pc_ratio_df = pd.read_html(str(table))[3]
        df = df.append(pc_ratio_df, ignore_index = True)
    
    # 日期欄位資料型態從文字轉為時間
    for row in range(df.shape[0]):
        date2 = df.iloc[row,0].split('/')
        df.iloc[row, 0] = datetime(int(date2[0]), int(date2[1]), int(date2[2]))
    
    # dataframe 按照日期排序
    df.sort_values("日期", inplace=True, ignore_index = True)
    
    # 設定 dataframe 的欄位名稱
    df.columns = ['日期', '賣權成交量', '買權成交量', '買賣權成交量比率%', '賣權未平倉量', '買權未平倉量', '買賣權未平倉量比率%']
    return df

df_pc['買賣權未平倉量比率差%'] = df_pc['買賣權未平倉量比率%'] - 100

def get_tw_market_data(start_year, start_month, end_year, end_month):
    start_date = str(date(start_year, start_month, 1))
    end_date = str(date(end_year, end_month, 1))
    month_list = pd.date_range(start_date, end_date, freq='MS').strftime("%Y%m%d").tolist()
    
    df = pd.DataFrame()
    for month in month_list:
        url = "https://www.twse.com.tw/exchangeReport/FMTQIK?response=json&date="+ month
        res = r.get(url)
        stock_json = res.json()
        stock_df = pd.DataFrame.from_dict(stock_json['data'])
        df = df.append(stock_df, ignore_index = True)
        
    # 資料轉型
    for col in [0, 1, 2, 3, 4]:
        for row in range(df.shape[0]):
            # 把"日期"從字串(string)換成時間(datetime)，並將民國年換成西元年
            if col == 0:
                day = df.iloc[row,0].split('/')
                df.iloc[row, 0] = datetime(int(day[0]) + 1911, int(day[1]), int(day[2]))  
            # 把"開盤價", "最高價", "最低價", "收盤價"帶有逗號的字串(string)換成浮點數(float) 
            elif col != 0:
                df.iloc[row, col] = float(df.iloc[row,col].replace(',', ''))
    
    df.columns = ['日期', '成交股數', '成交金額', '成交筆數', '發行量股價加權指數', '漲跌點數']
    return df