Tiny decoder for Chromium Web Custom MIME Data Format

decode_chromium_custom_mime.py

#!/usr/bin/env python3

# Refs:
# https://stackoverflow.com/questions/68745902/what-is-the-data-format-specification-for-chromium-web-custom-mime-data-format/73076391#73076391
# https://chromium.googlesource.com/chromium/src/+/refs/heads/main/ui/base/clipboard/custom_data_helper.cc

import sys
import json


def read_u16string(bs):
    length = int.from_bytes(bs[:4], "little")
    byte_length = length * 2
    if length % 2 != 0:
        byte_length += 2
    return bs[4 : 4 + length * 2].decode("utf-16le"), bs[4 + byte_length :]


def decode_chromium_web_custom(bs):
    # Format:
    # 0-3: UInt32LE length of data
    # 4-7: UInt32LE count of pairs
    # [count pairs of]
    #   U16String key
    #   U16String value
    # where U16String is:
    #   0-3: UInt32LE length of string
    #   4-4+length: UTF-16LE string
    data_len = int.from_bytes(bs[:4], "little")
    data = bs[4 : 4 + data_len]
    assert len(bs) - 4 == data_len
    count = int.from_bytes(data[:4], "little")
    data = data[4:]
    pairs = []
    for _ in range(count):
        key, data = read_u16string(data)
        value, data = read_u16string(data)
        pairs.append((key, value))
    return pairs


def encode_chromium_web_custom(pairs):
    data = b""
    for k, v in pairs:
        data += len(k).to_bytes(4, "little") + k.encode("utf-16le")
        if len(k) % 2 != 0:
            data += b"\0\0"
        data += len(v).to_bytes(4, "little") + v.encode("utf-16le")
        if len(v) % 2 != 0:
            data += b"\0\0"
    return len(data).to_bytes(4, "little") + len(pairs).to_bytes(4, "little") + data


def print_pairs(pairs):
    for k, v in pairs:
        if k.endswith("+json"):
            js = json.loads(v)
            print(f"{k}: {json.dumps(js, indent=2)}")
        else:
            print(f"{k}: {v}")
    print()

if __name__ == "__main__":
    pairs = decode_chromium_web_custom(open(sys.argv[1], "rb").read())
    print_pairs(pairs)