Update app.py

app.py

@@ -70,12 +70,11 @@ with st.sidebar:
 ''')
 
 
-st.subheader("YouTube Comments Sentiment Analysis", divider = "red")
-
+  
+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
 
@@ -105,94 +104,81 @@ if st.button("Sentiment Analysis", type="secondary"):
                 data = []
                 wait = WebDriverWait(driver, 30)
                 driver.get(url)
-                
+
+                placeholder = st.empty()
+                progress_bar = st.progress(0)
+
+                for item in range(150):
+                    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(3)
+                    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)
+
+                    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',
+                    )
                 st.warning("Failed to scrape reviews.")
-                update_url_count() # Correctly indented
+                update_url_count()
         else:
             st.warning("Please enter a URL.")
     else:
         st.warning(f"You have reached the maximum URL attempts ({max_attempts}).")
 
 st.write(f"URL pasted {st.session_state['url_count']} times.")
-
-
-    
-   
-    placeholder = st.empty()  # Create an empty placeholder for dynamic text
-    progress_bar = st.progress(0)  # Create a progress bar
-
-    for item in range(150):
-        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")  # Update placeholder text
-            progress_bar.progress((item + 1) / 150)  # Update progress bar
-            time.sleep(3)  # Increased sleep time for better loading
-        except Exception as e:
-            st.error(f"Exception during scrolling: {e}")
-            break
-
-    placeholder.text("Scrolling complete.") #show completion message.
-    progress_bar.empty() #remove progress bar.
-
-   
-    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  # Initialize unique user ID
-        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)
-
-            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',
-        )
-