W3C home > Mailing lists > Public > www-dom@w3.org > October to December 1998

Re: XML APIs

From: Arnaud Le Hors <lehors@w3.org>
Date: Tue, 03 Nov 1998 17:47:26 +0100
Message-ID: <363F339E.B3AA1DDE@w3.org>
To: DOM List <www-dom@w3.org>
John Cowan wrote:
> 
>         (leftSibling(rightSibling(node)) = node,
>                 unless node is a lastChild;
>         (rightSibling(leftSibling(node)) = node,
>                 unless node is a firstChild;
>         (firstChild(parent(node)) = node,
>                 if node is a firstChild;
>         (lastChild(parent(node)) = node,
>                 if node is a lastChild;
>         (rightSibling(firstChild(node)) = 2nd child of node
>                 provided there is one,
>         the algorithm "go to firstChild if there is one, otherwise
>         to rightSibling if there is one, otherwise to parent's
>         (grandparent's (great-grandparents (etc.))) rightSibling
>         until you reach the starting point" walks the descendants.
> 
> etc. etc. etc.  All of these are false under the DOM.

This not true. All of these _may_be_ false under the DOM. This is, they
will be false _if_ the DOM implementation, such as the browser, changes
the structure underneath as you're browsing through it.

This said, I also think this is not enough and I hope we can add some
locking mechanism in Level 2, so that you can make sure it is _always_
true.
-- 
Arnaud  Le Hors  -  W3C, DOM Activity Lead  - 
http://www.w3.org/People/Arnaud
Received on Tuesday, 3 November 1998 11:47:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 22 June 2012 06:13:46 GMT