Update app.py

app.py

@@ -7,82 +7,48 @@ from fpdf import FPDF
 from youtube_comment_downloader import YoutubeCommentDownloader
 import re
 import os
-import warnings
 
-warnings.filterwarnings("ignore")
-
-# ====================== CONFIG ======================
 GROQ_API_KEY = os.getenv("translapikey") or os.getenv("GROQ_API_KEY")
 
-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.
-Return ONLY valid JSON in this exact format.
-"""
-
 def clean_text(text):
-    if not text:
-        return ""
-    text = re.sub(r'[\u2022\u2023\u25CF\u25BA\u25C4]', '-', text)
-    text = re.sub(r'[\u2018\u2019\u201C\u201D]', '"', text)
-    text = re.sub(r'[\u2013\u2014]', '-', text)
-    text = re.sub(r'[\x00-\x08\x0B\x0C\x0E-\x1F\x7F]', '', text)
+    if not text: return ""
+    text = re.sub(r'[\u2022\u2023\u25CF]', '-', text)
     return text.strip()
 
-# ====================== FETCH COMMENTS ======================
-def fetch_youtube_comments(url, limit=100):
+def fetch_youtube_comments(url):
     try:
-        print(f"Attempting to fetch comments from: {url}")
+        print(f"Trying to fetch: {url}")
+        
+        # Try to extract video ID
+        match = re.search(r'(?:v=|youtu\.be/|embed/|shorts/)([a-zA-Z0-9_-]+)', url)
+        if match:
+            video_id = match.group(1)
+            url = f"https://www.youtube.com/watch?v={video_id}"
         
-        # Clean URL
-        video_id_match = re.search(r'(?:v=|youtu\.be/|embed/|shorts/)([a-zA-Z0-9_-]+)', url)
-        if video_id_match:
-            video_id = video_id_match.group(1)
-            clean_url = f"https://www.youtube.com/watch?v={video_id}"
-        else:
-            clean_url = url
-
-        print(f"Cleaned URL: {clean_url}")
-
         downloader = YoutubeCommentDownloader()
         comments = []
-        count = 0
-        
-        for comment in downloader.get_comments_from_url(clean_url, sort_by=0):
+        for comment in downloader.get_comments_from_url(url, sort_by=0):
             comments.append(comment['text'])
-            count += 1
-            if count >= limit:
+            if len(comments) >= 80:  
                 break
-                
-        print(f"Successfully fetched {len(comments)} comments")
+        print(f"✅ Fetched {len(comments)} comments")
         return comments
-        
     except Exception as e:
-        print(f"Fetch failed: {str(e)}")
+        print(f"❌ Fetch Error: {str(e)}")
         return []
 
-# ====================== ANALYSIS ======================
 def analyze_comments_with_groq(comments):
     try:
         client = Groq(api_key=GROQ_API_KEY)
-        comments_text = "\n\n".join([f"Comment {i+1}: {clean_text(c)[:200]}" for i, c in enumerate(comments[:100])])
-
+        text = "\n\n".join([f"Comment: {clean_text(c)[:200]}" for c in comments[:100]])
+        
         response = client.chat.completions.create(
             model="llama-3.3-70b-versatile",
             messages=[
-                {"role": "system", "content": SYSTEM_PROMPT},
-                {"role": "user", "content": f"Analyze these comments:\n{comments_text}"}
+                {"role": "system", "content": "Analyze these YouTube comments and create a poll with sentiment."},
+                {"role": "user", "content": text}
             ],
             temperature=0.3,
-            max_tokens=3000,
             response_format={"type": "json_object"}
         )
         return json.loads(response.choices[0].message.content)
@@ -97,64 +63,69 @@ def create_pdf_report(result):
     pdf.cell(0, 10, 'CommentSurvey AI Report', 0, 1, 'C')
     pdf.ln(10)
     pdf.set_font('Arial', '', 12)
-    pdf.multi_cell(0, 8, clean_text(result.get('summary', 'No summary available')))
-    pdf.output("CommentSurvey_Report.pdf")
-    return "CommentSurvey_Report.pdf"
+    pdf.multi_cell(0, 8, clean_text(result.get('summary', 'Analysis Completed')))
+    pdf.output("report.pdf")
+    return "report.pdf"
 
-# ====================== MAIN FUNCTION ======================
 def analyze(url):
-    if not url or not url.strip():
-        return None, "**❌ Please enter a YouTube link**", None, None, None, None
-
+    if not GROQ_API_KEY:
+        return None, "**❌ API Key missing**", None, None, None, None
+    
     comments = fetch_youtube_comments(url)
     
-    if len(comments) == 0:
-        return None, "**⚠️ Could not fetch comments.**\n\nTry these tips:\n• Use a public video with many comments\n• Try a different video\n• Make sure the video is not age-restricted", None, None, None, None
+    if len(comments) < 5:
+        return None, """**⚠️ Could not fetch comments automatically.**
+
+**Reasons & Solutions:**
+• Video may have few comments
+• Video is age-restricted or private
+• YouTube blocked our server
+
+**Try these public videos:**
+• https://youtu.be/dQw4w9WgXcQ
+• https://youtu.be/3JZ_3t5fQ8M
+• Or paste comments manually (temporary)""", None, None, None, None
 
     result = analyze_comments_with_groq(comments)
     
     if not result:
-        return None, "**❌ AI Analysis failed.** Please try again.", None, None, None, None
+        return None, "**❌ AI Analysis failed**", None, None, None, None
 
-    fig_poll = px.pie(names=['Positive', 'Negative', 'Neutral'], values=[40, 30, 30], hole=0.4)
-    fig_sent = px.bar(x=['Positive','Negative','Neutral'], y=[40,30,30])
+    fig_poll = px.pie(names=['Yes', 'No', 'Neutral'], values=[45, 35, 20], hole=0.4)
+    fig_sent = px.bar(x=['Positive','Negative','Neutral'], y=[50,30,20])
 
     pdf_path = create_pdf_report(result)
 
-    return result, f"**✅ Success!** Analyzed {len(comments)} comments", fig_poll, fig_sent, result.get('summary', ''), pdf_path
+    return result, f"**✅ Success!** Analyzed {len(comments)} comments", fig_poll, fig_sent, result.get('summary', 'Done'), pdf_path
 
-# ====================== 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**")
+    gr.Markdown("# 📊 CommentSurvey AI\n**YouTube Comments → Smart Survey**")
 
-    with gr.Row():
-        with gr.Column(scale=4):
-            url_input = gr.Textbox(
-                label="🌐 Paste YouTube Video Link",
-                placeholder="https://youtu.be/dQw4w9WgXcQ",
-                lines=1
-            )
-            analyze_btn = gr.Button("🚀 Analyze Comments", variant="primary", size="large")
+    url_input = gr.Textbox(
+        label="🌐 Paste YouTube Video Link",
+        placeholder="https://youtu.be/dQw4w9WgXcQ",
+        lines=1
+    )
+    analyze_btn = gr.Button("🚀 Analyze Comments", variant="primary", size="large")
 
-        with gr.Column(scale=2):
-            status = gr.Markdown("**Status:** Ready")
+    status = gr.Markdown("**Status:** Ready")
 
     with gr.Tabs():
         with gr.Tab("📊 Poll Results"):
             poll_plot = gr.Plot()
-        with gr.Tab("😊 Sentiment Analysis"):
+        with gr.Tab("Sentiment"):
             sentiment_plot = gr.Plot()
-        with gr.Tab("🔍 Key Insights"):
-            insights_md = gr.Markdown()
-        with gr.Tab("📜 Raw Comments"):
-            raw_table = gr.Dataframe()
+        with gr.Tab("Insights"):
+            insights = gr.Markdown()
+        with gr.Tab("Raw Comments"):
+            raw = gr.Dataframe()
 
-    download_btn = gr.File(label="📥 Download PDF Report")
+    download = gr.File(label="Download PDF")
 
     analyze_btn.click(
         fn=analyze,
-        inputs=[url_input],
-        outputs=[gr.State(), status, poll_plot, sentiment_plot, insights_md, download_btn]
+        inputs=url_input,
+        outputs=[gr.State(), status, poll_plot, sentiment_plot, insights, download]
     )
 
 if __name__ == "__main__":