A Quick Note About NOTATIONs

Before discussing attributes, it is necessary to briefly touch on one of the least-understood (and least-frequently used) parts of XML 1.0: the <!NOTATION> declaration. The NOTATION facility is essentially a way of associating an XML name identifier with a PUBLIC ID and/or a SYSTEM ID (a.k.a. URI). The specification does not dictate that an XML parser actually do anything with the IDs in a notation declaration, other than pass them to the client application.

Notations can be associated with external unparsed entities (such as binary image files) using the NDATA keyword in an entity declaration. They can also be associated with particular elements, by using the NOTATION attribute type (discussed below). A third use is to categorize processing instructions (by declaring a notation name that matches the processing instruction target name).

Besides these “official” uses, it is possible to declare notations for some other use within your application. Because information from all notation declarations is made available by an XML parser to the client application, an application is free to interpret them any way it wishes. But in most cases, a standardized approach would be preferable.