Extract the NFT contract address and token ID from a Reddit avatar URL.

decode_avatar_url.py

import base64
import re


def decode_avatar_url(avatar_url: str) -> tuple[str, int]:
    """Extract the NFT contract address and token ID from an avatar URL.

    >>> avatar_url = 'https://i.redd.it/snoovatar/avatars/nftv2_bmZ0X2VpcDE1NToxMzdfOWQ4NTQyZWYxMjM3OTMzYmFkYmU4NjcyOTFmNmMwNDM0YjhkMzE1Y18xNTY_rare_9c019f39-3606-43fa-b14e-7dfb2dc3c12a.png'
    >>> decode_avatar_url(avatar_url)
    ('0x9d8542ef1237933badbe867291f6c0434b8d315c', 156)
    """
    if not (match := re.search(r"/nftv2_([^_]+)_", avatar_url)):
        raise ValueError(f"Avatar URL is not an NFT avatar: {avatar_url!r}")

    base64_nft_info = match.group(1)
    # the base64 data is not padded, but python base64 lib needs padding with =
    pad = len(base64_nft_info) % 4 * "="
    nft_info = base64.b64decode(f"{base64_nft_info}{pad}").decode()
    # nft_info looks like 'nft_eip155:{CHAIN_ID}_{CONTRACT_ID}_{TOKEN_ID}'
    if not (match := re.match(r"nft_eip155:\d+_([a-fA-F0-9]{40})_(\d+)", nft_info)):
        raise ValueError(
            f"Avatar URL's NFT info data is not structured as expected: {nft_info!r}"
        )
    return f"0x{match.group(1)}", int(match.group(2))