Re: [xml] Cleaning the Web - Implementing HTML 5 parsing in libxml2

[Daniel Veillard <veillard redhat com>]

On Tue, Aug 26, 2008 at 09:36:37AM +0900, Karl Dubost wrote:
> Le 20 août 2008 à 23:34, Andi Sidwell a écrit :
> > FWIW, I've spent the summer working on a C HTML5 parser which is
> > approaching stability, called Hubbub[1].  It's about as half as fast  
> > as
> > libxml2 at parsing the HTML 5 spec with an O(1) treebuilder, and it's
> > fairly easy to bind to the libxml2 interfaces (and is being used in  
> > lieu
> > of the libxml2 HTML parser in a small Web browser, NetSurf[2], in the
> > development branch).  Note it's a) not buildable as a shared library  
> > or
> > b) had a formal release, but if someone wants an HTML5 parser in C,  
> > then
> > it's probably not a bad bet.
>
> excellent news. The HTML 5 Spec authorizes more than the usual event of 
> parsing by retrospectively modifying the tree (ala tidy), I wonder how 
> much it would require modification in libxml2 and if indeed it is a 
> better strategy to make an interface than directing including the code in 
> the library.

  Well, the big big difference is deployment, and maintaince !

Daniel

-- 
Daniel Veillard      | libxml Gnome XML XSLT toolkit  http://xmlsoft.org/
daniel veillard com  | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library  http://libvirt.org/