Two web titans, Amazon.com and Google.com, both recently launched experimental XML-based Web service APIs that alpha-geek developers can kick around to see how they might be used to incorporate these sites features into developers own Web sites. Amazon.coms API (available only to those in the sites Associates program) takes the first steps in opening up Amazon.coms huge product catalogs to others. Unfortunately, the API allows only searching by keyword or subject area; wed also like to have ISBN book-searching capabilities. Search results include common bibliographic information but not current prices or availability information.
The Amazon.com API is an unusual beast. Its not based on SOAP but instead has a URL-based calling syntax and returns a simple XML document instead of a SOAP message. Given the support that SOAP now has in development tools, a SOAP interface would allow easier site integration.
Google.coms API is based on SOAP and comes with bindings and sample code written in Java and C#. As with Amazon.coms API, callers need to register and pass in their ID with requests. Currently, the beta APIs are only for personal, noncommercial use (at home or at work) and are limited to 1,000 queries per day. Three Google.com features are exposed: a complete search interface, an API for retrieving old documents from the Google.com cache (a good way to make sure links to third-party content dont disappear) and an interface for Google.coms spelling checker (which is uniquely well-suited to correcting proper names).