Update app.py

app.py

@@ -13,6 +13,7 @@ import gradio as gr
 import torch
 import matplotlib.pyplot as plt
 from fpdf import FPDF
+import unicodedata
 
 # === Configuration ===
 persistent_dir = "/data/hf_cache"
@@ -223,17 +224,13 @@ Avoid repeating the same points multiple times.
     final_response = remove_duplicate_paragraphs(final_response)
     return final_response
 
-def generate_pdf_report_with_charts(summary: str, report_path: str):
-    import unicodedata
-
-    def clean_for_pdf(text):
-        # Remove emojis and any non-latin characters
-        return ''.join(c for c in text if unicodedata.category(c)[0] != 'So')
+def remove_non_ascii(text):
+    return unicodedata.normalize('NFKD', text).encode('ascii', 'ignore').decode('ascii')
 
+def generate_pdf_report_with_charts(summary: str, report_path: str):
     chart_dir = os.path.join(os.path.dirname(report_path), "charts")
     os.makedirs(chart_dir, exist_ok=True)
 
-    # Dummy chart
     chart_path = os.path.join(chart_dir, "summary_chart.png")
     categories = ['Diagnostics', 'Medications', 'Missed', 'Inconsistencies', 'Follow-up']
     values = [4, 2, 3, 1, 5]
@@ -244,7 +241,6 @@ def generate_pdf_report_with_charts(summary: str, report_path: str):
     plt.savefig(chart_path)
     plt.close()
 
-    # PDF report
     pdf_path = report_path.replace('.md', '.pdf')
     pdf = FPDF()
     pdf.add_page()
@@ -253,7 +249,8 @@ def generate_pdf_report_with_charts(summary: str, report_path: str):
     pdf.ln(5)
 
     for line in summary.split("\n"):
-        pdf.multi_cell(0, 10, txt=clean_for_pdf(line))
+        clean_line = remove_non_ascii(line)
+        pdf.multi_cell(0, 10, txt=clean_line)
 
     pdf.ln(10)
     pdf.image(chart_path, w=150)
@@ -264,31 +261,43 @@ def process_report(agent, file, messages: List[Dict[str, str]]) -> Tuple[List[Di
     if not file or not hasattr(file, "name"):
         messages.append({"role": "assistant", "content": "❌ Please upload a valid file."})
         return messages, None
+
     start_time = time.time()
     messages.append({"role": "user", "content": f"📂 Processing file: {os.path.basename(file.name)}"})
+
     try:
         extracted = extract_text(file.name)
         if not extracted:
             messages.append({"role": "assistant", "content": "❌ Could not extract text."})
             return messages, None
+
         chunks = split_text(extracted)
         batches = batch_chunks(chunks, batch_size=BATCH_SIZE)
         messages.append({"role": "assistant", "content": f"🔍 Split into {len(batches)} batches. Analyzing..."})
+
         batch_results = analyze_batches(agent, batches)
         valid = [res for res in batch_results if not res.startswith("❌")]
+
         if not valid:
             messages.append({"role": "assistant", "content": "❌ No valid batch outputs."})
             return messages, None
+
         summary = generate_final_summary(agent, "\n\n".join(valid))
+
         report_path = os.path.join(report_dir, f"report_{datetime.now().strftime('%Y%m%d_%H%M%S')}.md")
         with open(report_path, 'w', encoding='utf-8') as f:
             f.write(f"# Final Medical Report\n\n{summary}")
+
         pdf_path = generate_pdf_report_with_charts(summary, report_path)
+
         end_time = time.time()
         elapsed_time = end_time - start_time
+
         messages.append({"role": "assistant", "content": f"📊 **Final Report:**\n\n{summary}"})
         messages.append({"role": "assistant", "content": f"✅ Report generated in **{elapsed_time:.2f} seconds**.\n\n📥 PDF report ready: {os.path.basename(pdf_path)}"})
+
         return messages, pdf_path
+
     except Exception as e:
         messages.append({"role": "assistant", "content": f"❌ Error: {str(e)}"})
         return messages, None
@@ -302,22 +311,27 @@ def create_ui(agent):
         .gr-file, .gr-button { width: 100% !important; max-width: 400px; }
     """) as demo:
         gr.Markdown("""
-        <h2 style="text-align:center;">📄 CPS: Clinical Patient Support System</h2>
-        <p style="text-align:center;">Analyze and summarize unstructured medical files using AI (optimized for A100 GPU).</p>
+        <h2 style='text-align:center;'>📄 CPS: Clinical Patient Support System</h2>
+        <p style='text-align:center;'>Analyze and summarize unstructured medical files using AI (optimized for A100 GPU).</p>
         """)
+
         with gr.Column():
             chatbot = gr.Chatbot(label="🧠 CPS Assistant", height=480, type="messages")
             upload = gr.File(label="📂 Upload Medical File", file_types=[".xlsx", ".csv", ".pdf"])
             analyze = gr.Button("🧠 Analyze")
             download = gr.File(label="📥 Download Report", visible=False, interactive=False)
+
         state = gr.State(value=[])
+
         def handle_analysis(file, chat):
             messages, report_path = process_report(agent, file, chat)
             return messages, gr.update(visible=bool(report_path), value=report_path), messages
+
         analyze.click(fn=handle_analysis, inputs=[upload, state], outputs=[chatbot, download, state])
+
     return demo
 
 if __name__ == "__main__":
     agent = init_agent()
     ui = create_ui(agent)
-    ui.launch(server_name="0.0.0.0", server_port=7860, allowed_paths=["/data/hf_cache/reports"], share=False)
+    ui.launch(server_name="0.0.0.0", server_port=7860, allowed_paths=["/data/hf_cache/reports"], share=False)