- From: Mary Brady <mbrady@nist.gov>
- Date: Thu, 25 Oct 2001 12:20:57 -0400
- To: <www-dom-ts@w3.org>
----- Original Message ----- From: "Arnold, Curt" <Curt.Arnold@hyprotech.com> To: <www-dom-ts@w3.org> Sent: Thursday, October 25, 2001 12:14 PM Subject: RE: isSupported12.xml > In a perfect world all the following should return true (based on expectations from hasFeature and the recent errata on hasFeature, but not based on a detailed exegesis of the recommendation): > > isSupported("Core","2.0") > isSupported("XML","2.0") > isSupported("Core",null) > isSupported("XML",null) > isSupported("Core","") > isSupported("XML","") > // > // the last two are questionable, > // but the expectations would be that they are true > isSupported("Core","1.0") > isSupported("XML","1.0") > The spec indicates only 2.0 and not specified --> which maybe should be spelled out as either "" or null, as was done for hasFeature. > > Calling isSupported() with any other feature should simply not throw an exception. If a method throws an exception (and it isn't enclosed in a <assertDOMException> block), the test will fail, no > assertion needed. Since the main thing that you are checking is that isSupported() doesn't throw an exception on an unrecognized feature, the easiest way is just to put something like the following > at the bottom of the test: > The main thing we were checking was just the opposite -- that isSupported doesn't throw an exception with the known recognized features, as defined in the Conformance section of the recommendation. It is true that we need to stop somewhere, in that we can't specify additional XML names that others might create, but we can check for what the spec defines. --Mary > <!-- should not throw an exception, > can't assert an expectation on the return value --> > <isSupported var="isSupported" > feature='"!!!!InvalidFeature!!!!"' > version='"!!!!"'/> > >
Received on Thursday, 25 October 2001 12:32:28 UTC