Hi,

This looks nice.. We are already using JTidy for the HTML -> XHTML
conversion before we export to PDF.
A good way to do this is indeed as an XWiki plugin.
I would suggest 2 plugins:

1/ Importer
 It would be great if the importer have a generic part and a format
specific part. We could use the base code as the code to import from
other wiki formats.
 There is one case which is complex in importing is when you want to
handle links. You might need to read multiple files to convert links
from one format to another. For example if you have a set of HTML files
with relative links you might want to convert them to wiki links.
 An importer framework which allows to handle this would be great.

2/ Validator/Cleaner
 It would be great to be able to use each part separately and have some
options when calling it.
 Besides cross side scripting being able to allow/refuse tag lists would
be great and also allow/refuse velocity and groovy 

I guess Vincent would like to see the plugin outside of the core, but
I'll leave this to him.

Ludovic

Mac a écrit :
> Hey there,
>    I was thinking about building a XWiki plugin that will import
> (dynamically) another website, either whole or partial (A little XML
> parsing of it), besides reading the one page Plugin Doc Is there any
> thing else that I could use to help me speed up this development? Like
> XML parser/HtmlTidy or something that XWiki may already uses?
> 
> 
> Also,
>     While I am on this topic of html, since the Wiki will take
> Html/xml in the normal editing of the page.  I would be willing to
> write a Validator that would strip out dangerious html (Cross site
> scripting,...) I have already done this in a filter for a "Sudo wiki"
> I was building from scratch, but would be willing to re-write it in a
> way that would be helpful to the project, if someone could point me in
> the right direction where this type of Class would fit in to the project.
> 
> Thanks,
> 
> Cant wait to see beta 2
> 
> 
>