Update app.py

app.py

@@ -3,7 +3,7 @@ import json
 import plotly.express as px
 import pandas as pd
 from groq import Groq
-from fpdf2 import FPDF
+from fpdf import FPDF
 from youtube_comment_downloader import YoutubeCommentDownloader
 import re
 import os
@@ -12,14 +12,23 @@ import warnings
 warnings.filterwarnings("ignore")
 
 # ====================== CONFIG ======================
-GROQ_API_KEY = os.getenv("GROQ_API_KEY")
+#GROQ_API_KEY = os.getenv("translapikey") or os.getenv("GROQ_API_KEY")
+
+from google.colab import userdata
+GROQ_API_KEY = userdata.get('commentsurveyapi')
+
+
+if not GROQ_API_KEY:
+    print("❌ API Key not found!")
+else:
+    print("✅ API Key loaded successfully!")
 
 # ====================== SYSTEM PROMPT ======================
 SYSTEM_PROMPT = """
 You are an expert social media sentiment and poll analysis AI.
 Focus on Yes/No, Agree/Disagree, Support/Oppose, and sentiment.
 
-Handle English + Urdu well.
+Handle English + Urdu + Hindi + other languages well.
 Return ONLY valid JSON in this exact format:
 {
   "main_poll": {
@@ -45,6 +54,18 @@ Return ONLY valid JSON in this exact format:
 }
 """
 
+# ====================== TEXT CLEANING FOR UNICODE ======================
+def clean_text(text):
+    if not text:
+        return ""
+    # Replace problematic characters
+    text = re.sub(r'[\u2022\u2023\u25CF\u25BA\u25C4]', '-', text)  # bullets
+    text = re.sub(r'[\u2018\u2019\u201C\u201D]', '"', text)       # quotes
+    text = re.sub(r'[\u2013\u2014]', '-', text)                   # dashes
+    # Remove any remaining control characters
+    text = re.sub(r'[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]', '', text)
+    return text.strip()
+
 # ====================== HELPER FUNCTIONS ======================
 def extract_youtube_id(url):
     patterns = [
@@ -82,7 +103,9 @@ def fetch_youtube_comments(url, limit=150):
 def analyze_comments_with_groq(comments, post_context=""):
     try:
         client = Groq(api_key=GROQ_API_KEY)
-        comments_text = "\n\n".join([f"Comment {i+1}: {c[:250]}" for i, c in enumerate(comments[:130])])
+        # Clean comments before sending to Groq
+        cleaned_comments = [clean_text(c) for c in comments]
+        comments_text = "\n\n".join([f"Comment {i+1}: {c[:250]}" for i, c in enumerate(cleaned_comments[:130])])
 
         user_prompt = f"""
 Post Context: {post_context}
@@ -114,20 +137,20 @@ def create_pdf_report(analysis_result, poll_question):
     pdf.ln(10)
 
     pdf.set_font('Arial', 'B', 12)
-    pdf.cell(0, 10, f'Poll Question: {poll_question}', 0, 1, 'L')
+    pdf.cell(0, 10, clean_text(f'Poll Question: {poll_question}'), 0, 1, 'L')
     pdf.ln(5)
 
     pdf.set_font('Arial', 'B', 12)
     pdf.cell(0, 10, 'Summary:', 0, 1, 'L')
     pdf.set_font('Arial', '', 11)
-    pdf.multi_cell(0, 5, analysis_result.get('summary', 'No summary available.'))
+    pdf.multi_cell(0, 5, clean_text(analysis_result.get('summary', 'No summary available.')))
     pdf.ln(10)
 
     pdf.set_font('Arial', 'B', 12)
     pdf.cell(0, 10, 'Top Themes:', 0, 1, 'L')
     pdf.set_font('Arial', '', 11)
     for theme in analysis_result.get('top_themes', []):
-        pdf.cell(0, 7, f'• {theme}', 0, 1, 'L')
+        pdf.cell(0, 7, clean_text(f'• {theme}'), 0, 1, 'L')
 
     pdf.output("CommentSurvey_Report.pdf")
     return "CommentSurvey_Report.pdf"
@@ -136,7 +159,7 @@ def create_pdf_report(analysis_result, poll_question):
 def analyze(url):
     try:
         if not GROQ_API_KEY:
-            return None, "**❌ Groq API Key is missing!** Add it in Space Secrets.", None, None, None, None
+            return None, "**❌ Groq API Key is missing!**", None, None, None, None
 
         if not url or not url.strip():
             return None, "**❌ Please paste a YouTube URL**", None, None, None, None
@@ -180,7 +203,7 @@ def analyze(url):
     except Exception as e:
         return None, f"**❌ Error:** {str(e)}", None, None, None, None
 
-# ====================== GRADIO INTERFACE ======================
+# ====================== GRADIO UI ======================
 with gr.Blocks(title="CommentSurvey AI", theme=gr.themes.Soft()) as demo:
     gr.Markdown("# 📊 CommentSurvey AI\n**Turn YouTube Comments into Smart Surveys & Insights**")