- From: <keshlam@us.ibm.com>
- Date: Sun, 11 Jun 2000 23:32:21 -0400
- cc: www-dom@w3.org
These questions may already have been answered, but... >1) Will it be acceptable to replace the standard data members of a DOM interface (like >nodename) with get and set mechanism A DOM has no "standard data members". All DOM properties are accessed via setter and getter methods. Remember, the DOM is _only_ an API; there is no assurance that the structure of the data behind that API bears any resemblence to the tree-of-nodes view presented by the DOM. >2) Will it be acceptable to have one or more non-standard data members in the private >section of the implementation of the DOM interfaces? See above; data members are outside the scope of the DOM spec. The DOM only defines the behavior of the DOM API. If your implementation has other features as well, that's fine as long as the DOM APIs behave as described in the spec. (It's not uncommon for DOM implementations to add a few custom features. Of course any code which uses those features sacrifices portability.) >3) What do we mean by readonly declaration of a DOM specification data member See above. A readonly attribute has a getter method but no setter. >4) Also I am making the assumption that the name of the implementation of DOM >interface i.e. in the above case CDomNode (class name) can be non-standard. The abstract DOM interface should use the standard names. The class which inherits/implements that interface may use any name that's appropriate, as long as it can also be manipulated via the standardized name. (You may want to take a look at some of the existing Java and C++ implementations for examples. Or, perhaps, adapt one of those rather than creating your own from scratch.) ______________________________________ Joe Kesselman / IBM Research
Received on Monday, 12 June 2000 09:07:51 UTC