A phantomjs script to extract the list of kindle books from amazon.co.uk
phantomjs --ssl-protocol=any amazon.js [-latest] e-mail password
Two files:
- booklist.txt - a JSON stringified representation of the items
- booklist.csv - a CSV version of the relevant fields from the page
The optional -latest flag only extracts up to the latest 1000 books.
This code is quite fragile. It relies on an internal web service from the 'Manage Devices ...' page the web service only returns the first 1000 entries regardless of the batch size used.
Because the API only returns 1000 entries, it reads the list 6 ways. Note if you have over 2000 books you will not get a full listing - has been tested on a library with 3500+ books. The book list is queried by decending then ascending date, then title, then author. Duplicates are removed.
The Phantomjs browser needs to pretend to be an interactive browser (the userAgent string is from Chrome and may get out of date). If the user agent is not faked then amazon does not generate the relevant cookies.
** The Devices page can hang the browser. **
Need to use ssl-protocol flag on phantomjs as the default is not accepted by amazon at this time.
** Calls the UK site - not sure if this works for non-UK owners. **
The program sticks copious output on the console and outputs a number of screen shots to PNG
Happy hunting