Re: SEARCH draft 06 submitted

In a message dated 2/10/2004 9:37:34 AM Pacific Standard Time,  
julian.reschke@gmx.de writes:

FengAndy@aol.com wrote:

> Julian:
>  
> I  wonder why WebDAV Search is NOT based on XPath/XQuery.
>  
>  Conceptaully, a WebDAV storage contains metadata (property+folder) and  
> data blobs. WebDAV metadata could be easily modeled as XML.  Therefore, 
> XQuery/XPath could be applied for search.
>   
> Any comments?

Yes.

The DAV:basicsearch grammar is  *meant* to be simple, and it should map 
as directly as possible to a  simple database table.

Of course XPath would be a very interesting  *additional* query grammar 
for SEARCH, and I'm aware of at least two  experimental implementations. 
But I think it's too early to standardize on  that -- let's get the 
simpler things done first.

A DAV:xpath  grammar would belong into a separate spec. Anybody  volunteering?

Julian

-- 
<green/>bytes GmbH --  http://www.greenbytes.de -- tel:+492512807760

There are many XPath implementation available. For simiplicity,  we could 
limit to a subset of XPath grammar. By adopting XPath or its subset, it  should 
make WebDAV more easily adopted.
In future, if MSFT's longhorn becomes reality, search will  be key for 
WebDAV. Have any of you looked into Longhorn schema  definition?  It offers many 
many metadata on files/documents/media,  quite an opposite of current Windows FS. 
To support such an environment, WebDAV  overall may need to be extended.
BTW, is there anyone implementing WebDAV SEARCH?
Andy