use polars::prelude::*;

/// The standard metrics provided by JPMaQS (`value`, `grading`, `eop_lag`, `mop_lag`).
pub const DEFAULT_JPMAQS_METRICS: [&str; 4] = ["value", "grading", "eop_lag", "mop_lag"];

/// The required columns for a Quantamental DataFrame.
pub const QDF_INDEX_COLUMNS: [&str; 3] = ["real_date", "cid", "xcat"];


/// Check if a DataFrame is a quantamental DataFrame.
/// A standard Quantamental DataFrame has the following columns:
/// - `real_date`: Date column as a date type
/// - `cid`: Column of cross-sectional identifiers
/// - `xcat`: Column of extended categories
///
/// Additionally, the DataFrame should have atleast 1 more column.
/// Typically, this is one (or more) of the default JPMaQS metics.
pub fn is_quantamental_dataframe(df: &DataFrame) -> bool {
    let columns = df
        .get_column_names()
        .iter()
        .map(|s| s.as_str())
        .collect::<Vec<&str>>();
    let has_idx_columns = QDF_INDEX_COLUMNS.iter().all(|col| columns.contains(col));
    if !has_idx_columns {
        return false;
    }
    let real_date_col = df.select(["real_date"]);
    match real_date_col {
        Ok(_) => {}
        Err(_) => return false,
    };

    let is_date_dtype = real_date_col
        .unwrap()
        .dtypes()
        .iter()
        .all(|dtype| dtype == &DataType::Date);

    if !is_date_dtype {
        return false;
    }

    let cid_col = df.select(["cid"]);
    match cid_col {
        Ok(_) => {}
        Err(_) => return false,
    };

    let xcat_col = df.select(["xcat"]);
    match xcat_col {
        Ok(_) => {}
        Err(_) => return false,
    };

    // has atleast 1 more column
    let has_other_columns = columns.len() > 3;
    if !has_other_columns {
        return false;
    }

    return true;
}