If last scraped order in orders.json is more recent than end flag, than start date is more recent as end date and it crashes.

Example:
orders.json last date 2019

python -m scraping scrape --email [email protected] --password password --start 2015 --end 2017

--> start becomes 2019 and is therefore greater than end

If a item has no price we try to get that price via the details page by simply looking for the string behind "Summe:". This results in:

• not checking for discounts which are subtracted afterwards
• setting the full order price for a item when having multiple items in an order is very wrong
• setting the price as string instead of float is inconsistent with normal prices being floats

This field contains the aggregated shipping costs for all items inside that order.

Dash/Plotly produces a visually more appealing result

The current evaluation methods are static and could get pepped by some interaction features as setting the viewed time frame dynamically.

Plotly looks interesting:
https://plot.ly/python/

If one order contains multiple items they get grouped by the seller. For different sellers curently only each first item is detected.

I've no idea what could cause that...
Seem to appear more often when scraping everything instead of just a limited range.

• audiobook purchases are making problems
• adding balance to your account

one order can contain multiple items which isn't supported yet

Sometimes the price field for single item doesn't contain the price but some string as "Blitzangebote". Solution could be

1. setting price to 0.0, if no reconstruction is possible
2. setting price to total order price, for orders with only one item
3. reconstruct prize from subtracting prices of other items from total order price, remainder is this items price

Scraping the category for later grouping/evaluation would be useful. Requires going through each item individually and therefore is probably better to do it just as an option.

As of March 20th, Amazon.com (U.S.) no longer supports downloading the CSV files. Would you be willing to extend this to support Amazon.com?