app.py

import re
import tweepy
import pandas as pd
import gradio as gr
import itertools
import collections
from collections import Counter
import numpy as np
from transformers import pipeline
classifier = pipeline('sentiment-analysis')
summarizer= pipeline("summarization", max_length=10)
#hashtag_phrase ="#datascience"
#recent_tweet_count_you_want =100
def search_hashtag1(hashtag_phrase,recent_tweet_count_you_want):
  #hashtag_phrase=input("Enter hashtahg")
  consumer_key="30GAxNeTfZuPL5SfNhFBodmRF" 
  consumer_secret="C6O64nP0XjtwaAnXYL9zCcDZKEIP2iL1yVdlsNJtwLiZ5AEEBs"
  access_token="1246523558563471360-WrbCqO8phqjIzx393mrfOSKvDFPmey"
  access_token_secret="u7B6yX6ZyTa5ph7xkCFnbzyuD9jbuHHJNL0Y4S7mdZb1J"
  auth = tweepy.OAuthHandler(consumer_key, consumer_secret)
  auth.set_access_token(access_token, access_token_secret)
  api = tweepy.API(auth)
  fname = '_'.join(re.findall(r"#(\w+)", hashtag_phrase))
  data_frame=pd.DataFrame(columns={"timestamp"})
  timestamp=[]
  tweet_text=[]
  user_name=[]
  user_id=[]
  for tweet in tweepy.Cursor(api.search_tweets, q=hashtag_phrase+' -filter:retweets',lang="en", tweet_mode='extended').items(recent_tweet_count_you_want):
    timestamp1=tweet.created_at
    timestamp.append(timestamp1)
    #tweet_text1=tweet.full_text.replace('\n',' ').encode('utf-8')
    tweet_text1=tweet.full_text
    tweet_text.append(tweet_text1)
    user_name1=tweet.user.screen_name.encode('utf-8')
    user_name.append(user_name1)
    user_id1=tweet.id
    user_id.append(user_id1)
  data2=pd.DataFrame(timestamp,columns={"timestamp"})
  data1=pd.DataFrame(tweet_text,columns={"tweet_text"})
  data3=pd.DataFrame(user_name,columns={"user_name"})
  data4=pd.concat([data1,data2],axis=1)
  data5=pd.concat([data4,data3],axis=1)
  data7=pd.DataFrame(user_id,columns={"user_id"})
  data6=pd.concat([data5,data7],axis=1)
  tweet_list=data6.tweet_text.to_list()
  p = [i for i in classifier(tweet_list)]
  q=[p[i]['label'] for i in range(len(p))]
  data10=pd.DataFrame(q,columns={"sentiment"})
  data_tweet_final=pd.concat([data6,data10],axis=1)
  p_summarize_label = [i for i in summarizer(tweet_list)]
  q_summarize=[p_summarize_label[i]['summary_text'] for i in range(len(p_summarize_label))]
  data_summarize=pd.DataFrame(q_summarize,columns={"summarized_tweets"})
  data_tweet_summarize_final=pd.concat([data_tweet_final,data_summarize],axis=1)
  data_tweet_summarize_final.to_csv("tweet_data2.csv")
  #data6.to_csv("tweet_data1.csv")
  #data6=data5.head(10)
  return data_tweet_summarize_final
iface = gr.Interface(
    search_hashtag1,inputs=["text","number"],
    outputs="dataframe",
    examples=[["#datascience",5],["#valentine's day",10],["#pushpa",15],["#budget",20],["#sharktankindia",30]],
    theme="seafoam",
    title='Sakil Tweetlib6 App',
    description="You can extract tweets based on Hashtag.e.g. Please enter #datascience. The app extracts tweets based on the hashtag and the number of tweet count you want.")
iface.launch(inline=False)