Create app.py

app.py

@@ -0,0 +1,55 @@
+import gradio as gr
+from docling.datamodel.base_models import InputFormat
+from docling.datamodel.pipeline_options import PdfPipelineOptions
+from docling.document_converter import DocumentConverter, PdfFormatOption
+from docling_core.types.doc import ImageRefMode
+from pathlib import Path
+import os
+
+# Define the document converter
+pipeline_options = PdfPipelineOptions()
+pipeline_options.do_ocr = True
+pipeline_options.do_table_structure = True
+pipeline_options.table_structure_options.do_cell_matching = True
+pipeline_options.generate_picture_images = True
+
+doc_converter = DocumentConverter(
+    format_options={
+        InputFormat.PDF: PdfFormatOption(pipeline_options=pipeline_options)
+    }
+)
+
+def convert_to_markdown(file):
+    # Save the uploaded file to a temporary path
+    input_path = Path(file.name)
+    file.save(str(input_path))
+
+    # Convert the document
+    result = doc_converter.convert(str(input_path))
+
+    # Prepare output directory
+    output_dir = Path("output")
+    output_dir.mkdir(exist_ok=True)
+
+    # Save result as markdown
+    doc_filename = result.input.file.stem
+    md_filename = output_dir / f"{doc_filename}-with-images.md"
+    result.document.save_as_markdown(md_filename, image_mode=ImageRefMode.REFERENCED)
+
+    # Load the markdown content
+    with open(md_filename, 'r', encoding='utf-8') as f:
+        markdown_content = f.read()
+
+    return markdown_content
+
+# Create Gradio interface
+iface = gr.Interface(
+    fn=convert_to_markdown,
+    inputs=gr.File(label="Upload your document"),
+    outputs="markdown",
+    title="Document to Markdown Converter",
+    description="Upload a document (e.g., PDF, DOCX, PPTX) and get its Markdown version."
+)
+
+if __name__ == "__main__":
+    iface.launch(share=True)