Amazon ECS offers its affiliates a programmatic way of
fetching product information and building their own web sites using all the
creativity and deliver the traffic to its website.
Browse Nodes These are nothing but what most others call as product
categories. However, there is perhaps a reason why Amazon calls them Browse
Nodes. Amazon combines product categories with other types of groupings (such
as all books by a certain author, all products within a particular contry) and
provides a unified navigational interface. Amazon has millions of products and
hence, providing their entire catalog as a
data feed does not make
sense. However, the total number of browse nodes itself is probably no more
than a million and that too providing just the parent browse node id, browse node id and description as a data feed in a zipped format is not that big. Yet,
Amazon chose to provide the interface to the browse nodes only through their
ECS Web Services. So, I have a simple perl script that can
recursively download all the browse nodes starting from a given browse node. This script
is optimized to issue fewer number of service requests and also ensures that the
Amazon AWS Terms and Conditions are met (no more than one request per second).
LWP::UserAgent and XML::XPath are required to run this script.