Amazon Invoice Automation (Private Repository)

Log into Amazon

Log into Amazon using the following base snippet. Store data in session which you can use for future queries. Check response content to determine if its a successful login.

#!/usr/bin/env python3
import requests

session = requests.Session()
data = {'email':'email_id@gmail.com', 'password':'password'}
header={'User-Agent' : 'Mozilla/5.0'}

response = session.post('https://www.amazon.com/gp/sign-in.html', data, headers=header)
return response.content

Scrape Order Details Page for Order Numbers

Go to page

After successful login, go to order details page https://www.amazon.com/gp/css/order-history.

By default it will load order details for last 6 months which is sufficient for monthly automated invoice preparation. If required more, you can change some query parameters easily to suit your needs.

Count number of orders

Scrape Invoice links

Next find out all anchor tags with Invoice text inside them as well as a link that points to https://www.amazon.com/gp/css/summary/print.html/ref=oh_aui_pi_o00_?orderID=123-1234567-1234567, where order number is given by 123-1234567-1234567. In the link, base url (//www.amazon.com) might be omitted; so look for /gp/css/summary/print.html/ref=oh_aui_pi_o00_.

One sample link looks like: <a class="a-link-normal" href="/gp/css/summary/print.html/ref=oh_aui_pi_o00_?ie=UTF8&amp;orderID=123-1234567-1234567">Invoice</a>

Look into multiple pages

Number of orders might be more than 10 and hence we need to deal with pagination issues.

This is the url when I click page 2: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_1_2?ie=UTF8&orderFilter=months-6&search=&startIndex=10

This is the url for page 3: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_2_3?ie=UTF8&orderFilter=months-6&search=&startIndex=20

Explicit click to page 1: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_3_1?ie=UTF8&orderFilter=months-6&search=&startIndex=0

Clicking on next from page 1 to 2: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_1_2?ie=UTF8&orderFilter=months-6&search=&startIndex=10

Clicking on next from page 2 to 3: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_2_3?ie=UTF8&orderFilter=months-6&search=&startIndex=20

Clicking on previous from page 3 to 2: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_3_2?ie=UTF8&orderFilter=months-6&search=&startIndex=10

Clicking on previous from page 2 to 1: https://www.amazon.com/gp/your-account/order-history/ref=oh_aui_pagination_2_1?ie=UTF8&orderFilter=months-6&search=&startIndex=0

Scrape invoices (order numbers and corresponding links) till you hit the total invoice number found on first order history page all while discarding duplicate entries.

Go to order invoice page

Next, using the already stored requests.Session go to the invoice urls one by one. Sample url: https://www.amazon.com/gp/css/summary/print.html/ref=oh_aui_pi_o00_?orderID=123-1234567-1234567

invoice_url = 'https://www.amazon.com/gp/css/summary/print.html/ref=oh_aui_pi_o00_?orderID=123-1234567-1234567'
response = session.post(invoice_url, data, headers=header)

return response.content

Print the Invoices

You already have the response content or, the html content returned by amazon for that particular order number. Simply use that with the help of pdfkit to save file as a pdf with possibly the order number as part of filename.

import pdfkit

order_id = '123-1234567-1234567'
pdf_file = 'invoice_' + str(order_id) + '.pdf'
html_content = response.content

# pdfkit.from_url('http://amazon.com/...', pdf_file)
# pdfkit.from_file('test.html', pdf_file)
pdfkit.from_string(html_content, pdf_file)