SWECR API

This a proposal of a basic SWECR API which can be found here: http://swecr.googlecode.com/svn/trunk/org.swecr.api/.

The main principles of this API:

• This API is based on the Adapter pattern
• Each aspect of the API's functionality is clearly separated in an appropriate service (adapter)
• The API should be minimalistic. To perform an operation there is only one way to do it; no redundant methods or classes.
• This API should NOT be convenient for the end users; it should be simple to implement. All user-friendly operations should be implemented as a separate classes/libraries on top of this unique API.
• The API should provide the possibility to extend existing functionalities or implement it partially without modification of the API. (For this reason the Adapter pattern is used)

Some functional requirements:

• Each node is identified by an unique identifier. This id is not an URI. The mapping id <=> URI is performed in a separate service. This approach greatly simplifies resource re-naming, mapping of one Id to multiple URIs/URLs and provides the possibility of an efficient caching.
• This API should work with big binary data as well as with simple structured information (as strnigs, ints, ...)

Right now there is a Sesame2-based implementation of this API.