from pathlib import Path
from typing import Dict, Iterator, Optional, Tuple

from serato_doctor.models.database import DatabaseTrack, SeratoDatabase


TEXT_FIELDS = {
    b"tsng": "title",
    b"tart": "artist",
    b"talb": "album",
    b"tgen": "genre",
}


def iter_records(data: bytes) -> Iterator[Tuple[bytes, bytes]]:
    """Yield complete big-endian tag-length-value records."""

    offset = 0
    while offset + 8 <= len(data):
        tag = data[offset : offset + 4]
        length = int.from_bytes(data[offset + 4 : offset + 8], "big")
        payload_start = offset + 8
        payload_end = payload_start + length
        if payload_end > len(data):
            break
        yield tag, data[payload_start:payload_end]
        offset = payload_end


def decode_text(payload: bytes) -> Optional[str]:
    value = payload.decode("utf-16-be", errors="ignore").strip("\x00").strip()
    return value or None


def normalize_database_path(value: str) -> Path:
    if value.startswith(("Users/", "Volumes/")):
        value = "/" + value
    return Path(value)


def parse_track(payload: bytes) -> Optional[DatabaseTrack]:
    fields: Dict[str, Optional[str]] = {}
    path = None
    for tag, value in iter_records(payload):
        if tag == b"pfil":
            decoded_path = decode_text(value)
            if decoded_path:
                path = normalize_database_path(decoded_path)
        elif tag in TEXT_FIELDS:
            fields[TEXT_FIELDS[tag]] = decode_text(value)
    if path is None:
        return None
    return DatabaseTrack(path=path, filename=path.name, **fields)


def parse_database(database_path: Path) -> SeratoDatabase:
    data = database_path.read_bytes()
    version = None
    tracks = []
    for tag, payload in iter_records(data):
        if tag == b"vrsn":
            version = decode_text(payload)
        elif tag == b"otrk":
            track = parse_track(payload)
            if track is not None:
                tracks.append(track)
    return SeratoDatabase(database_path, version, tuple(tracks))