Polars dataframe to SQL Server using pyodbc, without Pandas or SQLAlchemy dependencies

pl_to_sql.py

def pl_to_sql(df, table, replace=False):
    """
    to_sql() without Pandas
    Appends or overwrites a SQL Server table using data from a DataFrame

    Parameters:
        df (DataFrame): df used to create/append table
        table (str): Name of existing SQL Server table
        replace (bool): Truncate before insert
    Returns:
        Nothing
    """

    try:

        cs = f'DRIVER={{ODBC Driver 18 for SQL Server}};SERVER={SERVER};DATABASE={DB};UID={UID};PWD={PWD}'

        cols = [f'[{c}]' for c in df.columns]
        cols = ','.join(cols)

        vals = '?' * len(df.columns)
        vals = ','.join(vals)

        sql = 'INSERT INTO ' + table + ' (' + cols + ') VALUES (' + vals + ')'

        with pyodbc.connect(cs) as conn:
            cursor = conn.cursor()
            cursor.fast_executemany = True

            if replace:
                cursor.execute(f"TRUNCATE TABLE {table}")

            cursor.executemany(sql, df.rows(named=False))
            conn.commit()
            cursor.close()

    except Exception:
        logger.opt(exception=True).error("cursor.executemany")
        raise