import_invoice.py

import psycopg2
import csv
from dotenv import load_dotenv
import os
from datetime import datetime

# Load environment variables
load_dotenv('../.env')

# Database connection parameters
conn_params = {
    "host": os.getenv("DATABASE_HOST"),
    "port": os.getenv("LOCAL_DATABASE_PORT"),
    "database": "amzn",
    "user": os.getenv("DATABASE_USER"),
    "password": os.getenv("DATABASE_PASSWORD")
}

# Path to the CSV file
csv_file_path = "/Users/tim/code/amzn_dwn/amazon-invoice-downloader/downloads/2024/transactions-2024.csv"

# Helper function to convert date strings to datetime objects
def convert_to_date(value):
    try:
        return datetime.strptime(value, '%Y-%m-%d')
    except ValueError:
        return None
    
def clean_amount(amount_str):
    if amount_str:
        return float(amount_str.replace('$', '').replace(',', '').strip())
    return None

# Main import logic
try:
    conn = psycopg2.connect(**conn_params)
    conn.autocommit = True  # Enable autocommit
    print("Database connection successful")
    cur = conn.cursor()
    
    # Define the table structure
    create_table_query = """
    CREATE TABLE IF NOT EXISTS invoice_data (
        date DATE,
        amount NUMERIC,
        card_type VARCHAR(50),
        last_4_digits VARCHAR(4),
        order_id VARCHAR(20),
        order_type VARCHAR(50)
    );
    """
    cur.execute(create_table_query)

    with open(csv_file_path, 'r', newline='', encoding='utf-8') as f:
        reader = csv.DictReader(f)
        for row_number, row in enumerate(reader, start=1):
            # Convert and clean data
            date = convert_to_date(row['Date'])
            amount = clean_amount(row['Amount'])
            card_type = row['Card Type']
            last_4_digits = row['Last 4 Digits']
            order_id = row['Order ID']
            order_type = row['Order Type']

            # Insert into the database
            try:
                cur.execute("""
                    INSERT INTO invoice_data (
                        date, amount, card_type, last_4_digits, order_id, order_type
                    ) VALUES (%s, %s, %s, %s, %s, %s)
                """, (date, amount, card_type, last_4_digits, order_id, order_type))
            except Exception as e:
                print(f"Row {row_number} insertion error: {e}")

except Exception as e:
    print(f"Error connecting to the database: {e}")

finally:
    if 'cur' in locals() and cur:
        cur.close()
    if 'conn' in locals() and conn:
        conn.close()

invoice_schema.sql

CREATE TABLE public.invoice_data (
    date date,
    amount numeric,
    card_type character varying(50),
    last_4_digits character varying(4),
    order_id character varying(20),
    order_type character varying(50)
);


ALTER TABLE public.invoice_data OWNER TO tim;

CREATE VIEW public.all_invoice_order_data AS
 SELECT o.order_id,
    (o.ship_date)::date AS ship_date,
    sum(o.shipment_item_subtotal) AS total_shipment_item_subtotal,
    sum(o.shipment_item_subtotal_tax) AS total_shipment_item_subtotal_tax,
    sum(o.shipping_charge) AS total_shipping,
    sum(o.total_owed) AS total_charge,
    string_agg(o.product_name, ' | '::text) AS concatenated_product_names,
    i.amount AS invoice_amount,
    i.date AS invoice_date
   FROM (public.orders o
     LEFT JOIN public.invoice_data i ON ((o.order_id = (i.order_id)::text)))
  WHERE ((o.order_status = 'Closed'::text) AND (((o.ship_date)::date >= '2024-01-01'::date) AND ((o.ship_date)::date <= '2024-12-31'::date)))
  GROUP BY o.order_id, ((o.ship_date)::date), i.amount, i.date
  ORDER BY (sum(o.total_owed)) DESC;


ALTER VIEW public.all_invoice_order_data OWNER TO tim;