import re
from pathlib import Path


TM_ID_PATTERN = re.compile(r"([CS]\d{4})", re.IGNORECASE)


def get_batch_upload_entries(config):

    bidding_config = config["exchange_bidding"]
    upload_root = Path(bidding_config["upload_input_root"])
    manifest_path = upload_root / "upload_manifest.json"

    if manifest_path.exists():
        return _load_manifest_entries(manifest_path)

    return _load_entries_from_filenames(upload_root)


def _load_manifest_entries(manifest_path):

    import json

    with manifest_path.open("r", encoding="utf-8") as file:
        payload = json.load(file)

    entries = []
    for item in payload.get("entries", []):
        entries.append(
            {
                "tm_id": item["tm_id"],
                "tac_id": item["tac_id"],
                "upload_path": str((manifest_path.parent / item["file_name"]).resolve()),
                "file_name": item["file_name"],
            }
        )
    return entries


def _load_entries_from_filenames(upload_root):

    entries = []
    for path in sorted(upload_root.iterdir()):
        if not path.is_file() or path.name.startswith("."):
            continue

        tac_id = path.stem.upper()
        tm_id = _derive_tm_id_from_tac(tac_id)
        entries.append(
            {
                "tm_id": tm_id,
                "tac_id": tac_id,
                "upload_path": str(path.resolve()),
                "file_name": path.name,
            }
        )

    if not entries:
        raise FileNotFoundError(f"No upload files were found in {upload_root}")

    return entries


def _derive_tm_id_from_tac(tac_id):

    match = TM_ID_PATTERN.search(tac_id)
    if not match:
        raise ValueError(
            f"Could not derive TM ID from TAC/file name '{tac_id}'. "
            "Use a manifest file for custom mapping."
        )
    return match.group(1).upper()