W3C home > Mailing lists > Public > www-webdav-dasl@w3.org > October to December 2003

Re: SEARCH by last path segment, Was: SEARCH for displayname

From: Elias Sinderson <elias@cse.ucsc.edu>
Date: Thu, 20 Nov 2003 04:04:45 -0800
Message-ID: <3FBCADDD.5060400@cse.ucsc.edu>
To: www-webdav-dasl@w3.org

Julian Reschke wrote:
> Elias Sinderson wrote:
>> What needs to be clarified is the relative ordering of evaluating 
>> include and exclude elements of a DASL request. [...]
> 
> [...] I think the rules we're defining do not depend on evaluation order. 

If order of evaluation doesn't matter, then the following would produce 
the same results:
all AND included AND NOT excluded
all AND NOT excluded AND included
This, however is clearly not the case as shown below.

Consider the scenario in which we have a collection, /A, containing 
resources /A/index.html, and /A/foo.pdf, both authored by Phil A. Novel. 
Further, let us suppose that we are evaluating a SEARCH request for 
resources authored by Mr. Novel and which specifies the scope as /A. In 
this case, the set 'all' (which satisfies the above constraints) is 
{/A/index.html, /A/foo.pdf}. For the sake of argument, let us assume 
that the values of the include-lastpathsegment and 
exclude-lastpathsegment elements are '%.pdf'.

Order of evaluation is important:
all AND included AND NOT excluded yields {/A/index.html}.
all AND NOT excluded AND included yields {/A/index.html, /A/foo.pdf}

Granted, this is a somewhat contrived example (albeit moreso for 
simplicity than otherwise), but I think it illustrates the idea that 
order of evaluation is important. The include and exclude sets don't 
need to be identical for evaluation order to matter, they only need to 
share some members in common. If you'd like I'll provide another 
illustrative example that is more realistic.


Cheers,
Elias
Received on Thursday, 20 November 2003 07:05:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 22 March 2009 03:38:10 GMT