How to append search keyword to twitter json data?

Question

I am doing twitter streaming data by kafka. I managed to stream the data and consume the twitter json. But now how do i create a pyspark dataframe containing the twitter data and the search keyword?

Below is how i write the kafka producer

I managed to create the dataframe of what data i want from the twitter object. But i don't know how to get the search keyword.

class StdOutListener(StreamListener):
def __init__(self, producer):
    self.producer_obj = producer

#on_status is activated whenever a tweet has been heard
def on_data(self, data):
    try:
        self.producer_obj.send("twitterstreamingdata", data.encode('utf-8'))
        print(data)
        return True
    except BaseException as e:
        print("Error on_data: %s" % str(e))
    return True

# When an error occurs
def on_error(self, status):
    print (status)
    return True

# When reach the rate limit
def on_limit(self, track):
    # Print rate limiting error
    print("Rate limited, continuing")
    # Continue mining tweets
    return True

# When timed out
def on_timeout(self):
    # Print timeout message
    print(sys.stderr, 'Timeout...')
    # Wait 10 seconds
    time.sleep(120)
    return True  # To continue listening

def on_disconnect(self, notice):
    #Called when twitter sends a disconnect notice
    return


if __name__ == '__main__':

spark = SparkSession \
    .builder \
    .appName("Kafka Producer Application") \
    .getOrCreate()

#This is the initialization of Kafka producer
producer = KafkaProducer(bootstrap_servers='xx.xxx.xxx.xxx:9092')

#This handles twitter auth and the conn to twitter streaming API
auth = OAuthHandler(consumer_key, consumer_secret)
auth.set_access_token(access_token, access_token_secret)
stream = Stream(auth, StdOutListener(producer))

print("Kafka Producer Application: ")

WORDS = input("Enter any words: ")
print ("Is this what you just said?", WORDS)
word = [u for u in WORDS.split(',')]
#This line filter twitter stream to capture data by keywords
stream.filter(track=word)

Answer 1

One way to resolve your problem it's changing StdOutListener class constructor to receive "keyword" parameter and add "keyword" to JSON in "on_data" function to send to Kafka

import json
import sys
import time

from kafka import KafkaProducer
from pyspark.sql import SparkSession
from tweepy import StreamListener, Stream, OAuthHandler


class StdOutListener(StreamListener):

    def __init__(self, producer: KafkaProducer = None, keyword=None):
        super(StreamListener, self).__init__()
        self.producer = producer
        self.keyword = keyword

    # on_status is activated whenever a tweet has been heard
    def on_data(self, data):
        try:
            data = json.loads(data)
            data['keyword'] = self.keyword
            data = json.dumps(data)
            self.producer.send("twitterstreamingdata", data.encode('utf-8'))
            return True
        except BaseException as e:
            print("Error on_data: %s" % str(e))
        return True

    # When an error occurs
    def on_error(self, status):
        print(status)
        return True

    # When reach the rate limit
    def on_limit(self, track):
        # Print rate limiting error
        print("Rate limited, continuing")
        # Continue mining tweets
        return True

    # When timed out
    def on_timeout(self):
        # Print timeout message
        print(sys.stderr, 'Timeout...')
        # Wait 10 seconds
        time.sleep(120)
        return True  # To continue listening

    def on_disconnect(self, notice):
        # Called when twitter sends a disconnect notice
        return


if __name__ == '__main__':
    CONSUMER_KEY = 'YOUR CONSUMER KEY'
    CONSUMER_SECRET = 'YOUR CONSUMER SECRET'
    ACCESS_TOKEN = 'YOUR ACCESS TOKEN'
    ACCESS_SECRET = 'YOUR ACCESS SECRET'

    print("Kafka Producer Application: ")
    words = input("Enter any words: ")
    print("Is this what you just said?", words)
    word = [u for u in words.split(',')]

    spark = SparkSession \
        .builder \
        .appName("Kafka Producer Application") \
        .getOrCreate()

    # This is the initialization of Kafka producer
    kafka_producer = KafkaProducer(bootstrap_servers='35.240.157.219:9092')
    # This handles twitter auth and the conn to twitter streaming API
    auth = OAuthHandler(CONSUMER_KEY, CONSUMER_SECRET)
    auth.set_access_token(ACCESS_TOKEN, ACCESS_SECRET)
    stream = Stream(auth, StdOutListener(producer=kafka_producer, keyword=word))
    stream.filter(track=word)

Hope it helps you!