Document-to-Podcase Snakemake adaptation

Snakefile

def download_document(document_url: str) -> bytes:
    """Download a document from the given URL.

    Args:
        document_url (str): The URL of the document to download.
    """
    import requests
    import logging

    logging.basicConfig(level=logging.INFO)
    logger = logging.getLogger(__name__)

    logger.info(f"Downloading document from {document_url}...")

    response = requests.get(document_url)
    response.raise_for_status()  # Raise an error for bad responses

    return response.content


rule download_document:
    output:
        output_path = "document.html"
    params:
        document_url = "https://lmsys.org/blog/2024-07-01-routellm/",
    run:
        content = download_document(params.document_url)
        with open(output.output_path, 'wb') as f:
            f.write(content)


rule convert_html_to_text:
    input:
        "document.html"
    output:
        "processed_document.txt"
    shell:
        """
        docker run --rm \
        -v $(pwd):/work \
        transformer-aarm64:v0.0.1 \
        --input /work/{input} \
        --output /work/{output}
        """


rule write_podcast_script:
    input:
        "processed_document.txt"
    output:
        "podcast_script.txt"
    shell:
        """
        docker run --rm \
        -v $(pwd):/work \
        text-to-script-aarm64:v0.0.1 \
        --input /work/{input} \
        --output /work/{output}
        """


rule generate_podcast:
    input:
        "podcast_script.txt"
    output:
        "podcast.wav"
    shell:
        """
        docker run --rm \
        -v $(pwd):/work \
        script_to_speech-aarm64:v0.0.1 \
        --input /work/{input} \
        --output /work/{output}
        """