i am trying to control the image and text alignment in my presentation
dynamically based on the user choice input.  the trick is to allow multiple
variations of image/text presentation on one page.

Page example:

paragraph 1 (user choice creates flag = left-right <xsl:if
test="body_image_placement='left-right'">)

	image flush left
	text right aligned to image


paragraph 2 (user choice creates flag = top-bottom <xsl:if
test="body_image_placement='top-bottom'">)

	image above text
	text below image

paragraph 3 (user choice creates flag = top-bottom <xsl:if
test="body_image_placement='top-bottom'">)

	no image
	text flush left

Rendered page:


paragraph 1		X	this is the text that appears
				right aligned to image



paragraph 2		X

			this is the text that appears below the
			image


paragraph 3		this is the text that apppears with no image


any suggestions?
tiffany_blake@xxxxxxx


-----Original Message-----
From: Michael Kay [mailto:mike@xxxxxxxxxxxx]
Sent: Monday, January 03, 2005 1:07 PM
To: 'Dimtre Novatchev'
Cc: xsl-list@xxxxxxxxxxxxxxxxxxxxxx
Subject: RE: [xsl] Computational complexity of accessing the Nth item in
a sequence and in a node-set


> So, if I have understood correctly, having
>
>    $sequence[last()]
>
> (and somehow more than one reference to $sequence)
> will guarantee that any further access to the items of $sequence will
> be performed in constant time?

No guarantees: but yes, that's what I would expect to happen.
>
> Cant this be pre-computed automatically by the XSLT processor?
> Something like computing a function with @memo-function="yes", but
> done by the XSLT processor?

I'm not clear what you mean.

Michael Kay
http://www.saxonica.com/