Re: [xml] redicting parts of trees

[Daniel Veillard <veillard redhat com>]

On Mon, May 16, 2005 at 01:59:34PM +0200, cazic gmx net wrote:
> I attached a first sketch for an adopt function. It tries the "oldNs"

  forgot the attachment I think :-)

> approach. It does not unlink the brach yet; I don't know if it should, or
> rather expect the branch to be unlinked already. The function is not tested,
> so just comments on its semantics please ;-)
> 
> I have the feeling that someone might object an addition to the code-base,
> which clearly leaves room for handling namespaces differently. So maybe this
> all ends up in a intermediate C layer for me; there won't be room for a "DOM
> wrapper helper" C file, will it?

  I won't object, unless it penalizes the comon case. For example if it forces
the use of an expensive callbck on each processed node, which would just be
an implementation mistake, not a principle error.

Daniel

-- 
Daniel Veillard      | Red Hat Desktop team http://redhat.com/
veillard redhat com  | libxml GNOME XML XSLT toolkit  http://xmlsoft.org/
http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/