[
Lists Home |
Date Index |
Thread Index
]
This is the kind of thing we have been discussing as a possible
requirement in the W3C XML Binary Characterization working group.
Some of us have also been working on solutions that provide the ability
to do this and several other features. My project is called Efficiency
Structured XML (esXML) at: http://www.esxml.org
sdw
Simon Kissane wrote:
>Hi,
>
>The standard text-based XML serialization is depth-first. So, for
>example, if the tree structure of my document is:
>Root
>|--- ElementA
>| |- ElementB1
>| | |- "Some text"
>| |- ElementB2
>|---- ElementC
> |-ElementD
>
>The standard depth-first XML serialization looks like:
><Root>
><ElementA><ElementB1>Some text</ElementB1><ElementB2/></ElementA>
><ElementC><ElementD /></ElementC>
></Root>
>
>But a breadth-first serialization could be something like:
>1 0 Root
>2 1 ElementA
>3 1 ElementC
>4 2 ElementB1
>5 2 ElementB2
>6 3 ElementD
>7 4 "Some text"
>where the first number is the number of the DOM node in breadth-first
>document order, and the second number is the number of the parent DOM
>node.
>
>Such a serialization might be useful in some cases. Suppose for
>example I have a huge XHTML document, with Javascript enabling
>drill-down to show/hide sections and subsections of the document, and
>initially all the sections are hidden. Using the standard depth-first
>serialization, I have to wait until the whole document loads to see
>the entirety of the top-level, whereas with a breadth-first
>serialization I could see the top-level immediately.
>
>More generally, one could imagine a transfer mechanism for XML which
>enabled the application to prioritise part of the DOM tree. So, for
>example, I click the "expand" button the last section of my huge HTML
>document, while it is loading. The renderering engine now needs to
>know what is under that node urgently. So it could send a request to
>the server saying "send DOM nodes under this node urgently, then keep
>on sending me the rest." The server would interrupt its normal
>transfer order to send those DOM nodes straight away, and then would
>resume sending the document as usual.
>
>Cheers
>Simon Kissane
>
>
>
>
>
>
--
swilliams@hpti.com http://www.hpti.com Per: sdw@lig.net http://sdw.st
Stephen D. Williams 703-724-0118W 703-995-0407Fax 20147-4622 AIM: sdw
|
