import base64
import subprocess
import yaml
import re
import os

# Load config.yaml for document properties
def load_config(config_path):
    with open(config_path, 'r', encoding='utf-8') as f:
        return yaml.safe_load(f)

if __name__ == "__main__":
    config = load_config('/app/data/config.yaml')


    # Step 1: Extract ```diagram blocks (base64-encoded SVG) and save as .svg files
    input_md = "/app/data/article.md"
    diagrams_dir = "/app/data/diagrams"
    os.makedirs(diagrams_dir, exist_ok=True)
    with open(input_md, 'r', encoding='utf-8') as f:
        md_content = f.read()

    # Replace diagram blocks with image references
    def replace_diagram_block(match):
        idx = replace_diagram_block.counter
        svg_content = base64.b64decode(match.group(1).strip())
        svg_path = os.path.join(diagrams_dir, f"diagram_{idx}.svg")
        with open(svg_path, 'wb') as sf:
            sf.write(svg_content)
        print(f"Extracted SVG diagram to {svg_path}")
        replace_diagram_block.counter += 1
        # Return markdown image reference
        return f'![Diagram {idx}](/app/data/diagrams/diagram_{idx}.svg)'
    replace_diagram_block.counter = 1

    new_md_content = re.sub(r'```diagram(.*?)```', replace_diagram_block, md_content, flags=re.DOTALL)

    # Save the modified markdown with image references
    temp_md = "/app/data/article_with_svgs.md"
    with open(temp_md, 'w', encoding='utf-8') as f:
        f.write(new_md_content)

    # Step 2: Convert Markdown to ODT using Pandoc
    output_odt = "/app/data/output.odt"
    subprocess.run([
        "pandoc",
        temp_md,
        "-o",
        output_odt
    ], check=True)

    # Step 2: Convert Markdown to ODT using Pandoc
    output_odt = "/app/data/output.odt"
    subprocess.run([
        "pandoc",
        input_md,
        "-o",
        output_odt
    ], check=True)

    print(f"Converted {input_md} to {output_odt}")