Since an understanding of urirefs is central to working with RDF, we'll take a moment to look at what makes a valid URI?the identifiers contained within a uriref and used to identify specific predicates.
Resources can be accessed with different protocols and using different syntaxes, such as using http:// to access a resource as a web page and ftp:// to access another resource using FTP. However, one thing each approach shares is the need to access a specific object given a unique name or identifier. URIs provide a common syntax for naming a resource regardless of the protocol used to access the resource. Best of all, the syntax can be extended to meet new needs and include new protocols.
URIs are related to URLs (Uniform Resource Locators) in that a URL is a specific instance of a URI scheme based on a known protocol, commonly the Hypertext Transfer Protocol (HTTP). URIs, and URLs for that matter, can include either a complete location or path to a resource or a partial or relative path. The URI can optionally include a fragment identifier, separated from the URI by a pound sign (#). In the following example, http://burningbird.net/articles/monsters3.htm is the URI and introduction is the fragment:
A URI is only an identifier. A specific protocol doesn't need to be specified, nor must the object identified physically exist on the Web?you don't have to specify a resolvable protocol such as http:// or ftp://, though you can if you like. Instead, you could use something as different as a UUID (Universally Unique Identifier) referencing a COM or other technology component that exists locally on the same machine or within a network of machines. In fact, a fundamental difference between a URL and a URI is that a URL is a location of an object, while a URI can function as a name or a location. URIs also differ from URNs (Uniform Resource Name) because URIs can refer to a location as well as a name, while URNs refer to globally unique names.
The RDF specification constrains all urirefs to be absolute or partial URIs. An absolute URI would be equivalent to the URL:
A relative URI is just as it sounds?relative to an absolute path. A relative reference to the Monsters article could be:
If a reference to the base location of the relative URI is not given, it's assumed to be base to the URI of the containing document. The use of URIs and the concepts of namespaces and QNames are discussed in more detail in Chapter 3.