RE: [xsl] RE: What are the differences between the different Saxon 9.2 versions and 9.1 (B) <----> 9.2 (HE) ?

Subject: RE: [xsl] RE: What are the differences between the different Saxon 9.2 versions and 9.1 (B) <----> 9.2 (HE) ?
From: Owen Rees <owen.rees@xxxxxx>
Date: Fri, 07 Aug 2009 12:54:30 +0100
--On 07 August 2009 11:58 +0100 Michael Kay wrote:

>  Michael Kay wrote:
>
>  http://snelson.org.uk/~jpcs/higher-order-functions.html

Thank you for the link.

Any reason why the subtyping relation is invariant for the
argument types instead of being more generally contravariant?

Sorry, I'm not in with the jargon. Can you explain the question?

A pointer to where in the text it says that the subtype relation is invariant would also be useful.


My first impression of section 2.3 is that the formula given for what "obviously matches" says that contravariant argument types are supported. (For those who do not know the jargon, it just means that the type relation for the arguments is in the opposite direction from the type relation for the result and the function as a whole.

--
Owen Rees; speaking personally, and not on behalf of HP.
========================================================
Hewlett-Packard Limited.   Registered No: 690597 England
Registered Office:  Cain Road, Bracknell, Berks RG12 1HN

Current Thread