- From: Joseph Kesselman <keshlam@us.ibm.com>
- Date: Mon, 11 Feb 2002 18:05:25 -0500
- To: Jonathan Watt <0jaw@qlink.queensu.ca>
- Cc: Boris Zbarsky <bzbarsky@MIT.EDU>, www-dom@w3.org
>Can I suggest that getElementsByName would be just as useful in other >things (as well as HTML). "Name" is not a defined concept that crosses all markup languages expected to be processed through the DOM. It's quite HTML-specific. We already have getElementsByNodeName. DOM Level 3 will presumably introduce the ability to retrieve elements by ID attribute, as part of the DTD/Schema support. A schema-specific set of APIs could support Keys as well. And of course you can always write a NodeFilter and plug it into a TreeWalker or NodeIterator -- which may not be as highly optimized as the above, but allows you the flexibility to define your own behaviors. Or -- assuming it survives into the REC and your DOM implements it -- try the XPath API. So there are lots of possible ways to address this use case in a more general and portable manner. I don't see a need to add getElementsByName, specifically, to the core DOM. Now, if the SVG DOM APIs wanted to add this method to _their_ APIs, because it was meaningful to them -- more power to them. ______________________________________ Joe Kesselman / IBM Research
Received on Monday, 11 February 2002 18:06:00 UTC