W3C home > Mailing lists > Public > www-dom-ts@w3.org > October 2001

Re: systemId within DocType Interface

From: Mary Brady <mbrady@nist.gov>
Date: Wed, 17 Oct 2001 12:54:53 -0400
Message-ID: <00f401c1572c$71b35440$293b0681@HAPPY>
To: <www-dom-ts@w3.org>
Given Phillipe's response, it would appear that the tests as
originally written are valid.  These tests -- one from DOM
Level 1 and another in DOM L2, capture the system id as a
string -- look backwards in the string for the first occurrence
of "/", and then strip off what would be the system id without
the baseURI.  This is done using the java String lastIndexOf
method, followed by a subString method.  In order to add
these test cases, we either need to add these language constructs
to the schema and language specific transformations, or we
could possibly add another assert element,
<assertURIequals>, with corresponding language specific
code to do the comparison.

Any thoughts?

--Mary


----- Original Message -----
From: "Philippe Le Hegaret" <plh@w3.org>
To: "Rick Rivello" <richard.rivello@nist.gov>
Cc: <www-dom@w3.org>
Sent: Wednesday, October 17, 2001 11:29 AM
Subject: Re: systemId within DocType Interface


> [moved to www-dom]
>
> Rick Rivello wrote:
> >
> > I'm going over the DOM Level 2 tests and have a question
> > about the systemId attribute within the DocumentType interface.
> >
> > Some of the parsers return the absolute URI while others return
> > just the file name.  Are both valid?  If not what should be returned?
>
> The DOM specification doesn't say anything about that. It is defined as
> a URI reference by the XML 1.0 specification. The Infoset doesn't say
anything
> either. So I guess both relative and absolute forms of the systemId are
valid
> in that case, even if DOM Level 2 doesn't give access to the baseURI.
>
> Philippe
>
>
Received on Wednesday, 17 October 2001 12:56:19 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 6 April 2009 12:58:45 GMT