app.py

import gradio as gr
import tempfile
import os
import json
import yaml
from pathlib import Path
from docling.datamodel.base_models import InputFormat
from docling.datamodel.pipeline_options import (
    PdfPipelineOptions,
    TesseractCliOcrOptions
)
from docling.document_converter import (
    DocumentConverter,
    PdfFormatOption,
    WordFormatOption
)
from docling.pipeline.standard_pdf_pipeline import StandardPdfPipeline
from docling.pipeline.simple_pipeline import SimplePipeline
from docling.backend.pypdfium2_backend import PyPdfiumDocumentBackend

# Initialize the document converter with advanced options
doc_converter = DocumentConverter(
    allowed_formats=[
        InputFormat.PDF,
        InputFormat.IMAGE,
        InputFormat.DOCX,
        InputFormat.HTML,
        InputFormat.PPTX,
        InputFormat.ASCIIDOC,
        InputFormat.MD,
        InputFormat.XLSX,
    ],
    format_options={
        InputFormat.PDF: PdfFormatOption(
            pipeline_cls=StandardPdfPipeline,
            backend=PyPdfiumDocumentBackend,
            pipeline_options=PdfPipelineOptions(
                do_ocr=True,
                ocr_options=TesseractCliOcrOptions(lang=["auto"])
            )
        ),
        InputFormat.DOCX: WordFormatOption(
            pipeline_cls=SimplePipeline
        ),
    }
)

def convert_document(input_file, output_format):
    try:
        # Get the input file path
        temp_dir = tempfile.mkdtemp()
        input_path = Path(input_file.name)
        
        # Generate output path
        output_filename = f"converted_document.{output_format.lower()}"
        output_path = Path(temp_dir) / output_filename
        
        # Convert the document
        result = doc_converter.convert(str(input_path))
        
        # Export to desired format
        if output_format.lower() == "html":
            with open(output_path, "w", encoding="utf-8") as f:
                f.write(result.document.export_to_html())
        elif output_format.lower() == "text":
            with open(output_path, "w", encoding="utf-8") as f:
                f.write(result.document.export_to_text())
        elif output_format.lower() in ["md", "markdown"]:
            with open(output_path, "w", encoding="utf-8") as f:
                f.write(result.document.export_to_markdown())
        elif output_format.lower() == "json":
            with open(output_path, "w", encoding="utf-8") as f:
                json.dump(result.document.export_to_dict(), f, indent=2)
        elif output_format.lower() == "yaml":
            with open(output_path, "w", encoding="utf-8") as f:
                yaml.safe_dump(result.document.export_to_dict(), f)
        elif output_format.lower() == "doctags":
            with open(output_path, "w", encoding="utf-8") as f:
                f.write(result.document.export_to_document_tokens())
        else:
            raise ValueError(f"Format de sortie non supporté: {output_format}")
        
        return str(output_path)
    except Exception as e:
        return f"Erreur lors de la conversion: {str(e)}"

# Define available formats
SUPPORTED_OUTPUT_FORMATS = [
    "HTML", "Markdown", "JSON", "Text", "YAML", "Doctags"
]

# Create the Gradio interface
with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue")) as demo:
    gr.Markdown(
        """
        # 📄 Document Converter Pro
        
        Convertissez facilement vos documents dans différents formats avec support OCR automatique.
        """
    )
    
    with gr.Row():
        with gr.Column(scale=1):
            input_file = gr.File(
                label="Télécharger un document",
                file_types=[
                    "pdf", "docx", "doc", "pptx", "xlsx",
                    "md", "html", "htm", "txt", "rtf",
                    "png", "jpg", "jpeg", "tiff", "bmp",
                    "adoc", "asciidoc"
                ],
            )
            output_format = gr.Dropdown(
                choices=SUPPORTED_OUTPUT_FORMATS,
                value="HTML",
                label="Convertir en",
                info="Sélectionnez le format de sortie"
            )
            convert_btn = gr.Button("Convertir le document", variant="primary")
        
        with gr.Column(scale=1):
            output = gr.File(label="Document converti")
            
    convert_btn.click(
        fn=convert_document,
        inputs=[input_file, output_format],
        outputs=output
    )
    
    gr.Markdown(
        """
        ### 📝 Instructions
        1. Téléchargez votre document
        2. Choisissez le format de sortie souhaité
        3. Cliquez sur "Convertir le document"
        
        ### ℹ️ Formats supportés
        
        #### Formats d'entrée
        - Documents Office : PDF, DOCX, XLSX, PPTX
        - Documents Web : HTML, XHTML
        - Documents texte : Markdown, AsciiDoc, TXT, RTF
        - Images : PNG, JPEG, TIFF, BMP
        
        #### Formats de sortie
        - HTML (avec support des images intégrées)
        - Markdown
        - Text (texte brut sans formatage)
        - JSON (sérialisation sans perte)
        - YAML
        - Doctags
        
        ### 🔍 Fonctionnalités
        - Détection automatique de la langue pour l'OCR
        - Support complet des tableaux
        - Extraction des images
        - Conversion multi-format
        """
    )

if __name__ == "__main__":
    demo.launch(share=False)