How to List Tar Archive Contents in Python

import tarfile
from pathlib import Path

def list_tar_contents(archive_path):
    """List all entries in a tar archive."""
    entries = []
    with tarfile.open(archive_path, "r") as tar:
        for member in tar.getmembers():
            entry_type = "dir" if member.isdir() else "file"
            entries.append(f"{entry_type:5s} {member.size:>10d} {member.name}")
    return entries


if __name__ == "__main__":
    # Create a small example tar archive to demonstrate
    import io
    import tarfile as tf

    sample_data = {
        "README.txt": b"Welcome to the archive\n",
        "config.json": b'{"version": 1}\n',
        "scripts/": None,
        "scripts/run.py": b'print("Hello")\n',
    }

    archive_path = Path("example.tar")
    with tf.open(archive_path, "w") as tar:
        for name, content in sample_data.items():
            info = tf.TarInfo(name)
            if content is None:
                info.type = tf.DIRTYPE
                tar.addfile(info)
            else:
                info.size = len(content)
                tar.addfile(info, io.BytesIO(content))

    # Now list the contents
    for line in list_tar_contents(archive_path):
        print(line)

    # Cleanup
    archive_path.unlink()