W3C home > Mailing lists > Public > public-ws-desc-eds@w3.org > September 2005

Re: WSDL components and properties stylesheets: properties attached to multiple components

From: Hugo Haas <hugo@w3.org>
Date: Tue, 27 Sep 2005 05:57:05 -0700
To: Arthur Ryman <ryman@ca.ibm.com>
Cc: public-ws-desc-eds@w3.org
Message-ID: <20050927125705.GB2255@w3.org>

Hi Arthur.

* Arthur Ryman <ryman@ca.ibm.com> [2005-09-23 16:38-0400]
> I don't like the idea of one component being treated differently, i.e. 
> with the others as dups.
> 
> There is no assumption now that a property name is unique to a component, 
> e.g. {name} appears in several components.
> 
> I think the safe way to handle this is to simply create a separate 
> <propdef> element for each occurence., e.g.
> 
>   <propdef comp="Binding">soap modules</propdef>
>   <propdef comp="Binding Operation">soap modules</propdef>
>   <propdup comp="Binding Message Reference">soap modules</propdef>
> 
> This will result in the property name appearing 3 times in the text, but 
> at least the tables will work.

I agree that it would be a good way to do it, but having "soap
modules" appear 3 times in the spec will look peculiar, so I was
looking for a mark-up specific way to deal with the problem.

> I think the cleanest XML design is to allow a comma-separated list for 
> @comp, e.g.
> 
> <propdef comp="Binding, Binding Operation, Binding Message Reference">soap 
> modules</propdef>
> 
> but that will require more work on the stylesheets (to process the list).

I thought to such a solution, but I think that it requires quite some
rework of the stylesheet:
- parsing and processing of the comma-separated list
- what value of @comp to require/allow for prop? probably all of them,
  which will again required extra processing

So I was proposing the propdup hack as a simple way to have the table
be accurate without reworking all the mechanism.

Cheers,

Hugo

-- 
Hugo Haas - W3C
mailto:hugo@w3.org - http://www.w3.org/People/Hugo/
Received on Tuesday, 27 September 2005 21:42:33 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Tuesday, 27 September 2005 21:42:34 GMT