I'm not sure if it's really doing that flattening or not, I'll take a closer
look.



On Sat, Apr 24, 2010 at 9:40 AM, P T Withington <[email protected]>wrote:

> This sounds half-baked and maybe obsolete to me.
>
> Not sure why we are putting everything into a namespace, you might have to
> ask Oliver for some history.
>
> And are you saying this transform is processing includes and flattening
> libraries?  How could that be, since the compiler later does the same thing?
>  Or is this happening after the compiler has expanded includes?
>
> On 2010-04-23, at 23:44, Henry Minsky wrote:
>
> > So I was trying to figure out why the LZX XML parser was adding a
> namespace
> > to every element, and
> > I eventually found that we apply an XSLT transform to the source file
> before
> > parsing it.
> >
> > That file is in lps/schema/preprocess.xsl
> >
> > Aside from slapping a namespace on every node that does not have one
> already
> > ,the file appears to have a transform to move top level
> > elements from library nodes to the canvas node.
> >
> > I'm wondering why we need to put a namespace onto all our nodes? Can't we
> > just leave them with the default (null)
> > namspace?
> >
> >
> >
> > And is it expensive to run the whole app source through
> > this XSL engine?  I   should measure the time spent in this transform on
> a
> > large application.
> >
> >
> >
> > --
> > Henry Minsky
> > Software Architect
> > [email protected]
>
>


-- 
Henry Minsky
Software Architect
[email protected]

Reply via email to