- From: Curt Arnold <carnold@houston.rr.com>
- Date: Sat, 5 Jan 2002 13:44:49 -0600
- To: "Sudhir Parikh" <sparikh@Resonate.com>
- Cc: <www-dom@w3.org>
Your question involves the specific behavior of one parser (my guess is Xerces-J) and so it would be more appropriate to ask on the xerces-j-user mailing list or another mailing list specific to the parser that you are using. The mechanics of creating a document object model from a XML document is outside the scope of current Level 2 DOM specs, though it is in the scope of the Level 3 Load and Save working draft. However, that draft is only experimentally implemented at this time. There are a couple of issues with your sample, first the URL for the DTD (C:\cs.dtd) is not a valid URL since backslashes are not allowed. See http://xml.apache.org/xerces2-j/faq-common.html#faq-1. Some parsers are forgiving when you use Windows UNC's, but that causes interoperability problems. Second, even if you aren't validating, some information from the DTD such as entity definitions and default attributes are required to properly build the DOM. See http://xml.apache.org/xerces2-j/faq-write.html#faq-5 If you want, you could register an EntityResolver with the parser (assuming Xerces-J or another DOM implementation layered on a SAX implementation) which will get called to resolve the reference to the external DTD. In the resolver, you could return an empty stream that would make it look like the file was empty which should allow parsing to continue.
Received on Saturday, 5 January 2002 14:47:42 UTC