def get_music_festivals_hack(paths: list[Path]) -> pd.DataFrame:
    """Shift music festival spends from problematic wslrs to other wslrs based on their
        national TV spend(as a proxy for propulation spread).
        Valid wholesalers to receive shifted spends are:
        1) Wholesalers with non-zero national TV spends which is used a weight.
        2) Wholesalers with non-zero music festival spends.
        3) Wholesalers who are not part of problematic wslrs

    Args:
        paths (list[Path]): Wholesaler data paths

    Returns:
        pd.DataFrame: Music festival spends to be updated.
    """
    # These wslrs have extreme spikes in music festivals
    problematic_wslrs = [
        "HENSLEY  CO PHOENIX (10261)",
        "HENSLEY  CO CHANDLER (10275)",
        "HENSLEY  CO PRESCOTT (10285)",
    ]
    # Use these brands to get approximate population spread of wholesalers
    top_brands = ["BUD", "MUL", "BDL"]
    # These are the weeks with extreme spikes in music festivals
    weeks_to_filter = [
        "2023-01-07",
        "2023-01-14",
        "2023-01-21",
        "2023-01-28",
        "2023-02-04",
        "2023-02-11",
        "2023-02-18",
        "2023-02-25",
    ]
    columns_to_read = ["week_date", "brand_code", "product_code", "league_Music_Festivals"]
    pop_spread = {}
    filtered_rows = []
    for i, current_path in enumerate(paths):
        wholesaler = current_path.parent.name.split("=")[-1]
        if wholesaler == "__HIVE_DEFAULT_PARTITION__":
            continue
        df = pd.read_parquet(current_path)
        df["week_date"] = pd.to_datetime(df["week_date"])
        df.rename(columns={"brand_cluster_code": "brand_code"}, inplace=True)
        if "product_code" not in df.columns:
            df["product_code"] = "NA"
        rows_to_fix = df[df.week_date.isin(weeks_to_filter)][columns_to_read]
        rows_to_fix["wholesaler"] = wholesaler
        filtered_rows.append(rows_to_fix)

        if wholesaler not in problematic_wslrs:
            # Use data since 2023 to get approximate population spread of wholesalers using national TV spend
            df = df[(df.week_date >= "2023-01-01") & (df.brand_code.isin(top_brands))]
            pop_spread[wholesaler] = df["vehicle_NATIONAL_TV"].sum()

    pop_spread = pd.DataFrame(list(pop_spread.items()), columns=["wholesaler", "weight"])
    pop_spread["weight"] = pop_spread["weight"] / pop_spread["weight"].sum()

    filtered_rows = pd.concat(filtered_rows)
    filtered_rows["issue_wlsr"] = filtered_rows["wholesaler"].isin(problematic_wslrs)
    filtered_rows = filtered_rows.merge(pop_spread, on="wholesaler", how="left")
    # Update music festival spends only when there is non-zero value.
    filtered_rows["weight"] = filtered_rows["weight"] * (filtered_rows["league_Music_Festivals"] > 0)

    music_festival_updates = []
    for week in filtered_rows.week_date.unique():
        for brand in filtered_rows.brand_code.unique():
            df = filtered_rows[(filtered_rows["week_date"] == week) & (filtered_rows["brand_code"] == brand)]
            # Spend to distribute for the week-brand - 2/3 of the spike's spend
            issue_spend = df[df.issue_wlsr]["league_Music_Festivals"].sum() * (2 / 3)
            df["weight"] = df["weight"] / df["weight"].sum()
            if df[~df.issue_wlsr]["weight"].isna().all() or issue_spend == 0:
                # Skip if no valid wslrs to distribute the spend
                continue
            spends_to_add = df["weight"] * issue_spend
            spends_to_remove = df["issue_wlsr"] * df["league_Music_Festivals"] * (-2 / 3)
            df["league_Music_Festivals"] = reduce(
                lambda x, y: x.add(y, fill_value=0), [df["league_Music_Festivals"], spends_to_add, spends_to_remove]
            )
            music_festival_updates.append(
                df[["week_date", "brand_code", "product_code", "wholesaler", "league_Music_Festivals"]]
            )
    if len(music_festival_updates) == 0:
        return None
    music_festival_updates = pd.concat(music_festival_updates).set_index(
        ["wholesaler", "product_code", "brand_code", "week_date"]
    )

    return music_festival_updates

def isolate_bdl_support_investments(output_df: pd.DataFrame, expected_spends: pd.DataFrame) -> pd.DataFrame:
    """Isolate other_cpnl_Incentives to other_cpnl_Bud_Light_Support and other_cpnl_Incentives"""
    output_df = output_df.copy()
    output_df["other_cpnl_Bud_Light_Support"] = 0.0
    spends_data = expected_spends.set_index("veh_names")
    bdl_support_frac = spends_data.loc["other_cpnl_Bud_Light_Support"] / (
        spends_data.loc["other_cpnl_Bud_Light_Support"] + spends_data.loc["other_cpnl_Incentives"]
    )
    bdl_support_frac.index = bdl_support_frac.index.astype(int)
    for year in bdl_support_frac.index:
        output_df.loc[output_df.index.get_level_values("week_date").year == year, "other_cpnl_Bud_Light_Support"] = (
            output_df.loc[output_df.index.get_level_values("week_date").year == year, "other_cpnl_Incentives"]
            * bdl_support_frac[year]
        )
        output_df.loc[output_df.index.get_level_values("week_date").year == year, "other_cpnl_Incentives"] *= (
            1 - bdl_support_frac[year]
        )
    return output_df