app.py

import urllib
import json
import os

from flask import Flask
from flask import request
from flask import make_response
from predictStocks import predictStocks
from twitter_analyze import twitter_analyze
from yahoo_finance import Share
from datetime import datetime, timedelta
import requests
# import mysql.connector

app = Flask(__name__)

# cnx = mysql.connector.connect(user=os.environ['JW_USERNAME'], password=os.environ['JW_KEY'], host=os.environ['JW_HOST'], database='xcqk05aruwtw0kew')

@app.route('/webhook', methods=['POST'])
def webhook():
    req = request.get_json(silent=True, force=True)

    print("Request:")
    print(json.dumps(req, indent=4))

    res = processRequest(req)

    res = json.dumps(res, indent=4)
    print(res)
    r = make_response(res)
    r.headers['Content-Type'] = 'application/json'
    print r
    return r

def processRequest(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")

    # logMessage(req)

    if req.get("result").get("action") == "CurrentPrice.price":   
        res = makeWebhookResult(getStockCurrentPrice(req), req, stock_symbol)
        return res
    elif req.get("result").get("action") == "Prediction.stockForecast":
        res = makeWebhookResult(getStockPrediction(req), req, stock_symbol)
        return res 
    elif req.get("result").get("action") == "Feelings.analyze":
        res = makeWebhookResult(getTwitterFeelings(req), req, stock_symbol)
        return res
    elif req.get("result").get("action") == "DividendDate.Date":
        res = makeWebhookResult(getStockDividendPayDate(req), req, stock_symbol)
        return res
    elif req.get("result").get("action") == "Stock.info":
        res = makeWebhookResult(getStockInfo(req), req, stock_symbol)
        return res
    elif req.get("result").get("action") == "Stock.historical":
        res = makeWebhookResult(getHistoricalData(req), req, stock_symbol)
        return res
    elif req.get("result").get("action") == "Decision.Classification":
        res = makeWebhookResult(getStockClassification(req), req, stock_symbol)
        return res 
    elif req.get("result").get("action") == "input.welcome":
        res = makeWebhookResult(getWelcome(req), req, stock_symbol)
        return res
    elif req.get("result").get("action") == "Visualize.chart":
        res = makeWebhookResult(getChartURL(req), req, stock_symbol)
        return res
    else:
        return {}

def getChartURL(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")
    chart_url = "https://www.etoro.com/markets/" + stock_symbol + "/chart"
    return chart_url

def logMessage(req):
    print "LOGGING!"
    originalRequest = req.get("originalRequest")
    source = ''

    if originalRequest != None:
        source = originalRequest.get("source")

    if source != 'facebook':
        print "not from facebook"
        return 

    data = originalRequest.get("data")
    time_stamp = data.get("timestamp")
    sender_id = data.get("sender").get("id")
    # recipient_id = data.get("recipient").get("id")
    message = data.get("message")
    text = message.get("text")

    # log incoming messagesw
    response = requests.post("http://api.botimize.io/messages?apikey=ZG2H9YHCZJQS9JTOTXXHL842QDGK5VHI", data={
      "platform": "facebook",
      "direction": "incoming",
      "raw": {
          "object":"page",
          "entry":[
            {
              "id":"986319728104533",
              "time":1458692752478,
              "messaging":[
                {
                  "sender":{
                    "id":sender_id
                  },
                  "recipient":{
                    "id":"986319728104533"
                  }
                }
              ]
            }
          ]
      }
    })
    print response
    print response.content 
    print "Success"


def getWelcome(req):
    response = 'Hi! I am here to help predict financial markets. My predictions are not 100% accurate!'
    return response

# analyze feelings intent
def getTwitterFeelings(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")
    if stock_symbol is None:
        return None

    twitter_analyzer = twitter_analyze()
    twitter_data = twitter_analyzer.analyze_feelings(stock_symbol)
    print 'Twitter data:'
    print twitter_data

    data = {}
    data['positive'] = twitter_data[0]
    data['negative'] = twitter_data[1]
    data['neutral'] = twitter_data[2]

    total = data['positive'] + data['negative'] + data['neutral']

    positive_percent = percentage(data['positive'], total)
    negative_percent = percentage(data['negative'], total)
    neutral_percent = percentage(data['neutral'], total)

    data_string = 'positive: ' + str(positive_percent) + '% negative: ' + str(negative_percent) + '% neutral: ' + str(neutral_percent) + '%'

    return data_string

# make percentage and round
def percentage(part, whole):
    return round(100 * float(part)/float(whole), 2)

# for intent prediction
def getStockPrediction(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")

    time = parameters.get("date-period")

    if stock_symbol is None:
        return None

    num_of_days = 3
    if time != '' and time is not None:
        num_of_days = extract_days(time)

    prediction = predictStocks()
    predicted_values = prediction.stocksRegression(stock_symbol, int(num_of_days))
    predicted_list = predicted_values.tolist()
    clean_list = cleanPrediction(predicted_list)

    return '\n'.join(str(v) for v in clean_list)

def cleanPrediction(list_prices):
    clean_list = []

    for price in list_prices:
        price_float = float(price[0])
        str_price =  '%.2f' % price_float
        clean_list.append(str_price)

    return clean_list

# invest or not
def getStockClassification(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")

    time = parameters.get("date-period")

    if stock_symbol is None:
        return None

    num_of_days = 14
    if time != '' and time is not None:
        num_of_days = extract_days(time)

    prediction = predictStocks()
    
    predicted_values = prediction.stocksNeuralNet(stock_symbol, int(num_of_days))
    predicted_decision = predicted_values.tolist()[-1][0]
    if time != '' and time is not None:
        return predicted_decision.lower() + ' (decision for next ' + str(num_of_days-1) + ' days)' 

    return predicted_decision.lower() + ' (decision for next two weeks)'

def extract_days(time):
    num_days = 3
    dates = time.split('/')

    if dates is not None:
        first = datetime.strptime(dates[0], "%Y-%m-%d").date()
        second = datetime.strptime(dates[1], "%Y-%m-%d").date()
        num_days = (second - first).days+1
    else:
        dates = time.split(' ')
        if dates is not None:
            if dates[0].isdigit():
                num_days = int(dates[0])

    return num_days

# intent current price
def getStockCurrentPrice(req):
    # db test
    # print 'Accessing database'
    # cursor = cnx.cursor(buffered = True)
    # str_call = 'SELECT * FROM USER_BASIC_INFO'
    # cursor.execute(str_call);

    # data = cursor.fetchone()
    # cnx.commit()
    # cursor.close()

    # print data
    # db test

    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")
    if stock_symbol is None:
        return None

    prediction = predictStocks()
    current_price = prediction.getCurrentPrice(stock_symbol)


    return str(current_price)


# intent dividend date
def getStockDividendPayDate(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")
    if stock_symbol is None:
        return None

    stock = Share(stock_symbol)
    pay_date = stock.get_dividend_pay_date()
    if pay_date is None:
        return 'No Dividend Date Avaliable'
    return str(pay_date)

def getStockInfo(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")
    if stock_symbol is None:
        return None

    stock = Share(stock_symbol)
    info = stock.get_info()
    return str(info)

# last 5 days data
def getHistoricalData(req):
    result = req.get("result")
    parameters = result.get("parameters")
    stock_symbol = parameters.get("stock_symbol")
    if stock_symbol is None:
        return None

    last_days = 5

    past_days_ago = datetime.now() - timedelta(days=last_days)
    past_days_ago_str = past_days_ago.strftime('%Y-%m-%d')

    now = datetime.now().date()
    now_str = now.strftime('%Y-%m-%d')

    stock = Share(stock_symbol)
    return str(stock.get_historical(past_days_ago_str, now_str))

# return to API.AI
def makeWebhookResult(data, req, stock_symbol):
    action = req.get("result").get("action")
    originalRequest1 = req.get("originalRequest")
    source = ''
    if originalRequest1 != None:
        source = originalRequest1.get("source")
    if action == "CurrentPrice.price":
        speech = "Current Price for the stock is $" + str(data)
        next_speech = "Predict " + stock_symbol
        # news_speech = "News for " + stock_symbol
        # news_url = "http://finance.yahoo.com/quote/" + stock_symbol
        chart_speech = "Chart for " + stock_symbol
        chart_url = "https://www.etoro.com/markets/" + stock_symbol + "/chart"
        feelings_speech = 'Feelings ' + stock_symbol
        if source == 'facebook':
            return {
                "speech": speech,
                "displayText": speech,
                "source": "apiai-wallstreetbot-webhook", 
                "data": {
                    "facebook": {
                      "attachment": {
                        "type": "template",
                        "payload": {
                                "template_type":"button",
                                "text":speech,
                                "buttons":[
                                  {
                                    "type":"web_url",
                                    "url":chart_url,
                                    "title":chart_speech,
                                    "webview_height_ratio": "compact"
                                  },
                                  {
                                    "type":"postback",
                                    "title":next_speech,
                                    "payload":next_speech
                                  },
                                  {
                                    "type":"postback",
                                    "title":feelings_speech,
                                    "payload":feelings_speech
                                  }
                                ]
                            }
                         }
                    }
                }
            }

    elif action == "Prediction.stockForecast":
        speech = "Predicted price for coming days: " + str(data)
    elif action == "Feelings.analyze":
        speech = "Feelings for " + stock_symbol + ": " + str(data)
    elif action == "Decision.Classification":
        speech = "I think we should " + str(data) + " " + stock_symbol 
        if source == 'facebook':
            return {
                "speech": speech,
                "displayText": speech,
                "source": "apiai-wallstreetbot-webhook", 
                "data": {
                    "facebook": {
                      "text":speech + '. Type away more questions!',
                        "quick_replies":[
                          {
                            "content_type":"text",
                            "title":"Need help?",
                            "payload":"Help"
                          }
                        ]
                    }
                }
            }

    elif action == "input.welcome":
        speech = str(data)
    elif action == "Visualize.chart":
        speech = 'Here is your chart:'
        chart_url = str(data)
        chart_speech = "Chart for " + stock_symbol

        if source == 'facebook':
            return {
                "speech": speech,
                "displayText": speech,
                "source": "apiai-wallstreetbot-webhook", 
                "data": {
                    "facebook": {
                      "attachment": {
                        "type": "template",
                        "payload": {
                                "template_type":"button",
                                "text":speech,
                                "buttons":[
                                  {
                                    "type":"web_url",
                                    "url":chart_url,
                                    "title":chart_speech,
                                    "webview_height_ratio": "compact"
                                  },
                                ]
                            }
                         }
                    }
                }
            }

    else:
        speech = str(data)

    print("Response:")
    print(speech)

    return {
        "speech": speech,
        "displayText": speech,
        "source": "apiai-wallstreetbot-webhook"
    }

    #gif example

    # Image example
    # "data": {
    #     "facebook": {
    #       "attachment": {
    #         "type": "image",
    #         "payload": {
    #         "url": "https://www.testclan.com/images/testbot/siege/weapons/assault-rifles.jpg"
    #          }
    #         }
    #       }
    #     }

    # quick reply template
  #   "message":{
  #   "text":"Pick a color:",
  #   "quick_replies":[
  #     {
  #       "content_type":"text",
  #       "title":"Red",
  #       "payload":"DEVELOPER_DEFINED_PAYLOAD_FOR_PICKING_RED"
  #     },
  #     {
  #       "content_type":"text",
  #       "title":"Green",
  #       "payload":"DEVELOPER_DEFINED_PAYLOAD_FOR_PICKING_GREEN"
  #     }
  #   ]
  # }

if __name__ == '__main__':
    port = int(os.getenv('PORT', 5000))

    print "Starting app on port %d" % port

    app.run(debug=False, port=port, host='0.0.0.0')