import pandas as pd
import argparse
from sqlalchemy import create_engine, text

# Function to read a parquet file and return a pandas dataframe
def read_parquet(file_path):
    try:
        # Reading the parquet file
        dataframe = pd.read_parquet(file_path, engine='pyarrow')
        return dataframe
    except Exception as e:
        print(f"Error reading parquet file: {e}")
        return None

# Function to convert dataframe to SQL
def dataframe_to_sql(df, table_name):
    # Using the Pandas and SQLAlchemy libraries to create a SQL string
    engine = create_engine('sqlite://', echo=False)
    df.to_sql(table_name, con=engine, if_exists='replace', index=False)

    # Extracting the SQL commands from the engine
    with engine.begin() as conn:
        create_table_statement = conn.execute(text(f"SELECT sql FROM sqlite_master WHERE name = '{table_name}'")).fetchone()[0]
        print(create_table_statement)

        rows = conn.execute(text(f"SELECT * FROM {table_name}")).fetchall()
        
        insert_statement = []
        for row in rows:
            placeholders = ', '.join(['?'] * len(row))
            insert_statement.append(f"({placeholders})".replace('?', '{}').format(*[repr(val) for val in row]))

        print(f"INSERT INTO {table_name} VALUES", (', ').join(insert_statement), ';')

# Main execution
if __name__ == "__main__":
    # Initialize argument parser
    parser = argparse.ArgumentParser(description="Read Apache Parquet file and convert its contents to SQL.")
    # Add argument for file path
    parser.add_argument('file_path', type=str, help='Path to the parquet file')
    parser.add_argument('table_name', type=str, help='Name of the SQL table')

    # Parse the arguments
    args = parser.parse_args()

    # Read the parquet file
    df = read_parquet(args.file_path)

    # Check if dataframe is not empty
    if df is not None:
        # Convert dataframe to SQL and print
        dataframe_to_sql(df, args.table_name)
    else:
        print("No data to process.")