import yfinance as yf
import os
import pandas as pd
import requests
import json
import datetime


#calculating columns how many days to expiration dates
def daysleft(inputdate):
  import datetime
  previous_date = datetime.datetime.strptime(inputdate, '%Y-%m-%d')
  today = datetime.datetime.today()
  ndays = (previous_date  - today).days +1
  return ndays

#get All Option Chain for a ticker, calculate additional columns and retun as a dataframe from Yahoo finance
def get_DF_optionchain(ticker):
    import requests_cache
    session = requests_cache.CachedSession('yfinance.cache')
    session.headers['User-agent'] = 'my-program/1.0'
    # ticker = yf.Ticker('msft', session=session)

    tkobj = yf.Ticker(ticker, session=session)
    exps = tkobj.options

    #   Current Price in Market Session
    #   lastprice = tkobj.info['currentPrice']
    lastprice = tkobj.history().iloc[-1]['Close']
    

  #base price for additional column 
    bias_price = 'mark'
#   bias_price = 'lastPrice'
    options = pd.DataFrame()
    for e in exps:
        try:
            print(e)
            opt = tkobj.option_chain(e)
            # opt_calls = pd.DataFrame().append(opt.calls)
            opt_calls = opt.calls
            opt_calls['Type'] = 'CALL'
            # opt_puts = pd.DataFrame().append(opt.puts)
            opt_puts = opt.puts
            opt_puts['Type'] = 'PUT'
            opt = opt_calls._append(opt_puts)
            opt['expirationDate'] = e
            opt['daysleft'] = daysleft(e)

            opt[['bid', 'ask', 'strike']] = opt[['bid', 'ask', 'strike']].apply(pd.to_numeric)
            opt['mark'] = (opt['bid'] + opt['ask']) / 2
            opt['pricepercent'] = 100 * opt['mark'] / lastprice

            opt[['bid', 'ask', 'strike']] = opt[['bid', 'ask', 'strike']].apply(pd.to_numeric)
            opt['mark'] = (opt['bid'] + opt['ask']) / 2

            opt['interinsicvalue'] = opt.apply(lambda row: abs(lastprice - row['strike']) if row['inTheMoney'] else 0, axis=1)
            opt['timevalue'] = opt.apply(lambda row: row[bias_price] - row['interinsicvalue'], axis=1)
            opt['breakevenprice'] = opt.apply(lambda row: row[bias_price]+ row['strike'] if row['Type']=='CALL' else row['strike'] - row['lastPrice'], axis=1)
            options = options._append(opt, ignore_index=True)    
        except:
            print('ERROR')
            pass
    return options



pathdatafolder = os.path.join('data' , datetime.datetime.today().strftime("%Y%m%d"))
if not os.path.exists(fr"/app/{pathdatafolder}"):
    os.mkdir(fr"/app/{pathdatafolder}")

tickers = ['TSLA', 'META', 'GOOG', 'IBM', 'MSFT','NKE','DLTR','DG']

for ticker in tickers:
    print(ticker)
    try:
        options = get_DF_optionchain(ticker)
        options = options.drop(['contractSize','currency'] , axis=1)
        options.to_csv(fr"/app/{pathdatafolder}/{ticker}.csv",index=False)
    except:
        pass


# df = pd.read_csv(r"\\PERSIA\New Volume\storage\premarket\america_2022-04-14.csv")

# df2 = df.query("`Market Capitalization`>10e9")

# import timeit
# start = timeit.default_timer()


# for index,row in df2.iterrows():
#     ticker = row.Ticker
#     print(ticker)
#     try:
#         options = get_DF_optionchain(ticker)
#         options.to_csv(fr"./{pathdatafolder}/{ticker}.csv")
#     except:
#         pass



# stop = timeit.default_timer()
# print('Time: ', stop - start)  




# import os
# import glob

# DFtotal = pd.DataFrame()
# for csvfile in glob.glob(r'./data/*.csv'):
#     DF = pd.read_csv(csvfile)
#     DF['Ticker'] = os.path.basename(csvfile).split('.')[0]
#     DFtotal = DFtotal.append(DF)
# DFtotal.shape
# DFtotal.columns


# DFtotal.sort_values(by='impliedVolatility')[-10:].T
# DFtotal.sort_values(by='impliedVolatility')[-10:].T


# DFtotal.query('volume>100').sort_values(by='impliedVolatility')[-10:].T
# DFtotal.query('daysleft>7 & daysleft<60 & not inTheMoney & Type=="CALL" & openInterest>1000').sort_values(by='pricepercent')[-5:].T

# dftest = DFtotal.query('daysleft>7 & daysleft<60 & inTheMoney!=1 & Type=="CALL" &openInterest>100').sort_values(by='pricepercent')
# dftest['lasttradedate']=dftest['lastTradeDate'].str.split(' ', 0, expand=False).str[0]
# dftest.to_csv('test.csv')
# # dftest['lasttradedate']=pd.to_datetime(dftest['lastTradeDate'])
# dftest['lasttradedate2']=pd.to_datetime(dftest['lasttradedate'],format='%Y-%m-%d')
# dftest['year'] = pd.DatetimeIndex(dftest['lasttradedate']).year
# dftest['month'] = pd.DatetimeIndex(dftest['lasttradedate']).month
# dftest['day'] = pd.DatetimeIndex(dftest['lasttradedate']).day
# dftest.info()
# dftest.query('year==2022 & month==4 & day==14').sort_values(by='pricepercent')
# [-5:].T

# # dftest.query('lasttradedate>2022-04-14')

# df = pd.read_csv(r"./{pathdatafolder}/AAPL.csv")
# df.columns
# df.sort_values(by='impliedVolatility')[-5:].T

# df.query('impliedVolatility>1').sort_values(by='openInterest')[-5:].T

# df.query('daysleft>7 & daysleft<60 & not inTheMoney & Type=="CALL" & openInterest>10000').sort_values(by='bid')[-5:].T