Update app.py

app.py

@@ -5,13 +5,14 @@ from transformers import pipeline
 from keybert import KeyBERT
 from fpdf import FPDF
 import os
+import re
 
 # --- SETUP ---
-openai.api_key = os.getenv("OPENAI_API_KEY")  # Set this in your HF Space Secrets
+openai.api_key = os.getenv("OPENAI_API_KEY")  # Set in HF Space Secrets
 
 summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
 kw_model = KeyBERT()
-FONT_PATH = "DejaVuSans.ttf"  # Upload this to your Space!
+FONT_PATH = "DejaVuSans.ttf"  # Must be uploaded to Space root!
 
 BRANDS = [
     "Apple", "Google", "Microsoft", "Amazon", "Coca-Cola", "Pepsi", "Samsung", "Nike", "Adidas",
@@ -19,7 +20,6 @@ BRANDS = [
     "Infosys", "Wipro", "Paytm", "Zomato", "Swiggy", "OLA", "Uber"
 ]
 
-# --- HELPERS ---
 def extract_brands(text):
     found = [brand for brand in BRANDS if brand.lower() in text.lower()]
     return found if found else ["None detected"]
@@ -34,11 +34,35 @@ def make_bullets(summary):
     bullets = [f"- {s.strip()}" for s in sentences if s.strip()]
     return "\n".join(bullets)
 
-# --- PDF REPORT CREATION ---
+def make_str(val):
+    try:
+        if val is None:
+            return ""
+        if isinstance(val, (bool, int, float)):
+            return str(val)
+        if isinstance(val, list):
+            return "\n".join([make_str(v) for v in val])
+        if isinstance(val, dict):
+            return str(val)
+        return str(val)
+    except Exception:
+        return ""
+
+def safe_multicell(pdf, text, w=0, h=8):
+    """Safely adds text to PDF, handling super long words."""
+    if not isinstance(text, str):
+        text = str(text)
+    # Split very long words (>80 chars) to avoid fpdf2 crash
+    def break_long_words(t, maxlen=80):
+        return re.sub(r'(\S{%d,})' % maxlen,
+                      lambda m: ' '.join([m.group(0)[i:i+maxlen] for i in range(0, len(m.group(0)), maxlen)]),
+                      t)
+    text = break_long_words(text)
+    pdf.multi_cell(w, h, text)
+
 def create_pdf_report(language, transcript, transcript_en, summary, brands, topics, key_takeaways):
     pdf = FPDF()
     pdf.add_page()
-    # Use DejaVu Unicode font (must be uploaded to Space root)
     pdf.add_font("DejaVu", style="", fname=FONT_PATH, uni=True)
     pdf.add_font("DejaVu", style="B", fname=FONT_PATH, uni=True)
     pdf.set_font("DejaVu", "B", 16)
@@ -47,42 +71,27 @@ def create_pdf_report(language, transcript, transcript_en, summary, brands, topi
     pdf.ln(5)
     pdf.cell(0, 10, f"Detected Language: {language}", ln=True)
     pdf.ln(5)
-    pdf.multi_cell(0, 8, "Original Transcript:\n" + (transcript or ""))
+    safe_multicell(pdf, "Original Transcript:\n" + (transcript or ""))
     pdf.ln(3)
-    pdf.multi_cell(0, 8, "English Transcript:\n" + (transcript_en or ""))
+    safe_multicell(pdf, "English Transcript:\n" + (transcript_en or ""))
     pdf.ln(3)
     pdf.set_font("DejaVu", "B", 12)
     pdf.cell(0, 10, "Brands Detected:", ln=True)
     pdf.set_font("DejaVu", size=12)
-    pdf.multi_cell(0, 8, ", ".join(brands))
+    safe_multicell(pdf, ", ".join(brands))
     pdf.set_font("DejaVu", "B", 12)
     pdf.cell(0, 10, "Key Topics:", ln=True)
     pdf.set_font("DejaVu", size=12)
-    pdf.multi_cell(0, 8, ", ".join(topics))
+    safe_multicell(pdf, ", ".join(topics))
     pdf.set_font("DejaVu", "B", 12)
     pdf.cell(0, 10, "Summary (Bulleted):", ln=True)
     pdf.set_font("DejaVu", size=12)
     for takeaway in key_takeaways.split('\n'):
-        pdf.multi_cell(0, 8, takeaway)
+        safe_multicell(pdf, takeaway)
     pdf_file = "/tmp/analysis_report.pdf"
     pdf.output(pdf_file)
     return pdf_file
 
-def make_str(val):
-    try:
-        if val is None:
-            return ""
-        if isinstance(val, (bool, int, float)):
-            return str(val)
-        if isinstance(val, list):
-            return "\n".join([make_str(v) for v in val])
-        if isinstance(val, dict):
-            return str(val)
-        return str(val)
-    except Exception:
-        return ""
-
-# --- MAIN APP FUNCTION ---
 def process_audio(audio_path):
     if not audio_path or not isinstance(audio_path, str):
         return ("No audio file provided.", "", "", "", "", "", None)
@@ -118,11 +127,9 @@ def process_audio(audio_path):
         summary = summary_obj[0]["summary_text"] if isinstance(summary_obj, list) and "summary_text" in summary_obj[0] else make_str(summary_obj)
     except Exception as e:
         summary = f"Error summarizing: {e}"
-    # New: Brands, topics, bullets
     brands = extract_brands(transcript_en)
     topics = extract_topics(transcript_en)
     key_takeaways = make_bullets(summary)
-    # PDF file generation
     pdf_file = create_pdf_report(lang_text, transcript, transcript_en, summary, brands, topics, key_takeaways)
     return (
         lang_text,
@@ -134,7 +141,6 @@ def process_audio(audio_path):
         pdf_file
     )
 
-# --- GRADIO INTERFACE ---
 iface = gr.Interface(
     fn=process_audio,
     inputs=gr.Audio(type="filepath", label="Upload MP3/WAV Audio"),
@@ -148,7 +154,7 @@ iface = gr.Interface(
         gr.File(label="Download PDF Report")
     ],
     title="Audio Transcript, Brand & Topic Analysis (OpenAI Whisper + Unicode PDF Download)",
-    description="Upload your audio file (MP3/WAV). Get full transcript, summary, brand and topic detection, and download results as PDF."
+    description="Upload your audio file (MP3/WAV). Get transcript, summary, brand & topic detection, and download PDF. Unicode (Indian language/emoji) supported."
 )
 
 iface.launch()