app.py

import streamlit as st 

from selenium import webdriver
from selenium.webdriver.common.by import By
from selenium.webdriver.chrome.options import Options

from selenium.webdriver.chrome.service import Service

import pandas as pd

from selenium.webdriver.common.keys import Keys

from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC
import time
import sys
from datetime import datetime


from webdriver_manager.chrome import ChromeDriverManager
from selenium.webdriver.chrome.service import Service as ChromeService

from webdriver_manager.core.os_manager import ChromeType

import re


import transformers
from transformers import DistilBertTokenizer, DistilBertForSequenceClassification
import io
import plotly.express as px
import zipfile
import torch 

import matplotlib.pyplot as plt
from wordcloud import WordCloud
from nltk.corpus import stopwords

import nltk
nltk.download('stopwords')



with st.sidebar:
    st.button("DEMO APP", type="primary")
   

    expander = st.expander("**Important notes on the YouTube Comments Sentiment Analysis App**")
    expander.write('''
    
    
    **How to Use**
    This app works with a YouTube URL.  Paste the URL and press the 'Sentiment Analysis' button to perform sentiment analysis on your YouTube Comments.
    
    
    **Usage Limits**
    You can perform sentiment analysis on YouTube Comments up to 5 times.
    
    
    **Subscription Management**
    This demo app offers a one-day subscription, expiring after 24 hours. If you are interested in building your own YouTube Comments Sentiment Analysis Web App, we invite you to explore our NLP Web App Store on our website. You can select your desired features, place your order, and we will deliver your custom app in five business days. If you wish to delete your Account with us, please contact us at [email protected]
    
    
    **Customization**
    To change the app's background color to white or black, click the three-dot menu on the right-hand side of your app, go to Settings and then Choose app theme, colors and fonts.
    
    
    **Charts**
    Hover to interact with and download the charts.
    
    
    **File Handling and Errors**
    For any errors or inquiries, please contact us at [email protected]
   
    
    
''')


st.subheader("YouTube Comments Sentiment Analysis", divider="red")
tokenizer = transformers.DistilBertTokenizer.from_pretrained("tabularisai/robust-sentiment-analysis")
model = transformers.DistilBertForSequenceClassification.from_pretrained("tabularisai/robust-sentiment-analysis")

if 'url_count' not in st.session_state:
    st.session_state['url_count'] = 0

max_attempts = 2

def update_url_count():
    st.session_state['url_count'] += 1

def clear_question():
    st.session_state["url"] = ""

url = st.text_input("Enter YouTube URL:", key="url")
st.button("Clear question", on_click=clear_question)

if st.button("Sentiment Analysis", type="secondary"):
    if st.session_state['url_count'] < max_attempts:
        if url:
            with st.spinner("Wait for it...", show_time=True):
                options = Options()
                options.add_argument("--headless")
                options.add_argument("--disable-gpu")
                options.add_argument("--no-sandbox")
                options.add_argument("--disable-dev-shm-usage")
                options.add_argument("--start-maximized")
                service = Service(ChromeDriverManager(chrome_type=ChromeType.CHROMIUM).install())
                driver = webdriver.Chrome(service=service, options=options)
                data = []
                wait = WebDriverWait(driver, 30)
                driver.get(url)

                placeholder = st.empty()
                progress_bar = st.progress(0)

                for item in range(30):
                    try:
                        body = WebDriverWait(driver, 30).until(EC.visibility_of_element_located((By.TAG_NAME, "body")))
                        body.send_keys(Keys.END)
                        placeholder.text(f"Scrolled {item + 1} times")
                        progress_bar.progress((item + 1) / 150)
                        time.sleep(0.5)
                    except Exception as e:
                        st.error(f"Exception during scrolling: {e}")
                        break

                placeholder.text("Scrolling complete.")
                progress_bar.empty()

                data = []
                try:
                    wait.until(EC.presence_of_element_located((By.CSS_SELECTOR, "#contents #contents")))
                    comments = driver.find_elements(By.CSS_SELECTOR, "#content #content-text")
                    user_id = 1
                    for comment in comments:
                        timestamp = datetime.now().strftime("%Y-%m-%d")
                        data.append({"User ID": user_id, "Comment": comment.text, "comment_date": timestamp})
                        user_id += 1
                    data = [dict(t) for t in {tuple(d.items()) for d in data}]
                except Exception as e:
                    st.error(f"Exception during comment extraction: {e}")
                driver.quit()
                df = pd.DataFrame(data, columns=["User ID", "Comment", "comment_date"])
                st.dataframe(df)

                if tokenizer and model:
                    inputs = tokenizer(df['Comment'].tolist(), return_tensors="pt", padding=True, truncation=True)
                    with torch.no_grad():
                        logits = model(**inputs).logits
                        predicted_probabilities = torch.nn.functional.softmax(logits, dim=-1)
                        predicted_labels = predicted_probabilities.argmax(dim=1)
                        results = []
                        for i, label in enumerate(predicted_labels):
                            results.append({'Review Number': i + 1, 'Sentiment': model.config.id2label[label.item()]})
                        sentiment_df = pd.DataFrame(results)

                    value_counts1 = sentiment_df['Sentiment'].value_counts().rename_axis('Sentiment').reset_index(name='count')
                    final_df = value_counts1
                    tab1, tab2 = st.tabs(["Pie Chart", "Bar Chart"])
                    with tab1:
                        fig1 = px.pie(final_df, values='count', names='Sentiment', hover_data=['count'], labels={'count': 'count'})
                        fig1.update_traces(textposition='inside', textinfo='percent+label')
                        st.plotly_chart(fig1)

                    result = pd.concat([df, sentiment_df], axis=1)
                    st.dataframe(result)

                    with tab2:
                        fig2 = px.bar(result, x="Sentiment", y="comment_date", color="Sentiment")
                        st.plotly_chart(fig2)

                    text = " ".join(review for review in df['Review'])
                    stopwords = set(stopwords.words('english'))
                    text = re.sub('[^A-Za-z]+', ' ', text)
                    words = text.split()
                    clean_text = [word for word in words if word.lower() not in stopwords]
                    clean_text = ' '.join(clean_text)
                    stopwords = set(stopwords.words('english'))
                    wc = WordCloud(width=800, height=400, background_color='white').generate(clean_text)
                    fig = plt.figure(figsize=(12,6))
                    plt.imshow(wc, interpolation='bilinear')
                    plt.axis('off')

                    st.pyplot(fig)


                    csv = result.to_csv(index=False)
                    st.download_button(
                        label="Download data as CSV",
                        data=csv,
                        file_name='Summary of the results.csv',
                        mime='text/csv',
                    )
                
        else:
            st.warning("Please enter a URL.")
    else:
        st.warning(f"You have reached the maximum URL attempts ({max_attempts}).")

if 'url_count' in st.session_state: #added if statement.
    st.write(f"URL pasted {st.session_state['url_count']} times.")