Update app.py

app.py

@@ -1,26 +1,34 @@
 import gradio as gr
 import subprocess
 import os
-import shutil
+import base64
+import uuid
 
-def convert_docx_to_pdf(docx_file):
-    if docx_file is None:
-        return None
+def process_conversion(file_input, base64_input):
+    temp_docx = f"input_{uuid.uuid4().hex}.docx"
+    output_dir = "outputs"
     
-    # Get the file path from the Gradio temp file object
-    input_path = docx_file.name
-    filename = os.path.basename(input_path)
-    output_dir = "output_pdfs"
-    
-    # Create output directory if it doesn't exist
     if not os.path.exists(output_dir):
         os.makedirs(output_dir)
-        
+
     try:
-        # Run LibreOffice headless conversion command
-        # --headless: Runs without a GUI
-        # --convert-to pdf: Specifies target format
-        # --outdir: Specifies where the result should go
+        # 1. Determine Input Source
+        if file_input is not None:
+            # User uploaded a file
+            input_path = file_input.name
+        elif base64_input and base64_input.strip():
+            # User provided a base64 string
+            # Clean base64 string (remove headers if present, e.g., "data:application/..." )
+            if "," in base64_input:
+                base64_input = base64_input.split(",")[1]
+            
+            with open(temp_docx, "wb") as f:
+                f.write(base64.b64decode(base64_input))
+            input_path = temp_docx
+        else:
+            return None, "Error: Please provide either a file or a Base64 string."
+
+        # 2. Run LibreOffice Conversion
         subprocess.run([
             'libreoffice', 
             '--headless', 
@@ -28,28 +36,54 @@ def convert_docx_to_pdf(docx_file):
             '--outdir', output_dir, 
             input_path
         ], check=True)
-        
-        # Determine the name of the generated PDF
+
+        # 3. Locate the Output PDF
+        # LibreOffice names the output the same as input but with .pdf
+        filename = os.path.basename(input_path)
         pdf_filename = os.path.splitext(filename)[0] + ".pdf"
         pdf_path = os.path.join(output_dir, pdf_filename)
-        
+
         if os.path.exists(pdf_path):
-            return pdf_path
-        else:
-            return "Error: PDF was not generated."
+            # Read PDF to return as Base64 string
+            with open(pdf_path, "rb") as f:
+                pdf_bytes = f.read()
+                pdf_base64 = base64.b64encode(pdf_bytes).decode('utf-8')
             
+            # Clean up temp docx if we created one
+            if os.path.exists(temp_docx):
+                os.remove(temp_docx)
+                
+            return pdf_path, pdf_base64
+        else:
+            return None, "Error: Conversion failed."
+
     except Exception as e:
-        return f"An error occurred: {str(e)}"
-
-# Define Gradio Interface
-demo = gr.Interface(
-    fn=convert_docx_to_pdf,
-    inputs=gr.File(label="Upload DOCX File", file_types=[".docx"]),
-    outputs=gr.File(label="Download PDF"),
-    title="DOCX to PDF Converter",
-    description="Upload a .docx file and convert it to .pdf using LibreOffice Headless. High accuracy conversion for Hugging Face Spaces.",
-    allow_flagging="never"
-)
+        return None, f"Error: {str(e)}"
+
+# Custom CSS for a better look
+css = """
+#container { max-width: 800px; margin: auto; }
+"""
+
+with gr.Blocks(css=css) as demo:
+    gr.Markdown("# 📄 DOCX to PDF Converter")
+    gr.Markdown("Upload a Word file **OR** paste a Base64 encoded string of the binary DOCX.")
+    
+    with gr.Row():
+        with gr.Column():
+            file_in = gr.File(label="Upload .docx", file_types=[".docx"])
+            text_in = gr.Textbox(label="Paste Base64 DOCX", placeholder="Base64 data here...", lines=5)
+            submit_btn = gr.Button("Convert to PDF", variant="primary")
+            
+        with gr.Column():
+            file_out = gr.File(label="Download PDF")
+            text_out = gr.Textbox(label="Result Base64 PDF", show_copy_button=True, lines=5)
+
+    submit_btn.click(
+        fn=process_conversion,
+        inputs=[file_in, text_in],
+        outputs=[file_out, text_out]
+    )
 
 if __name__ == "__main__":
     demo.launch()