Re: No-fail `document` Function Spec

On Tue, Mar 07, 2006 at 06:30:56PM -0700, Mike Brown wrote:
> The generate-id() equality requirement prevents us from getting too
> carried away with that, though...at least for a given evaluation of a
> stylesheet. As it stands, you *do* have a guarantee that 'document'
> represents / will return the same node-set for the same URL,
> throughout the stylesheet (for a given run).

Ok, then the current `res:document` spec almost behaves the same way,
because of its reference to the original `document` function.  In order
to extend the "same node-set guarantee", I think I need to say something
about the identity of the generated "result" content.  I'll work on
these modifications and try to submit a new draft soon.

There may also be an opportunity for some new functionality here.  What
do you think about the idea of adding a third parameter to the
`res:document` function representing an attempt number, such that the
`res:document` function is guaranteed to be the same for the combination
of the URI and the attempt number?  This attempt number would default to
1, so that by default the function provides the "old behavior".  But
then users could do things like the following:

  res:document($uri)
  res:document($uri, ., 2)

Where the second call is in effect a "second attempt", and could
retrieve a different document.  Further calls with an attempt number of
2 (and the same $uri) would be guaranteed to return the same result, but
this allows the user to increment the result number to try retrieving
the document again.

Thoughts?

Take care,

    John L. Clark

_______________________________________________
exslt mailing list
[email protected]
http://www.exslt.org/list

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.7 (GNU/Linux)

iD8DBQFEDyDdjkcskvOyl78RAgydAKDXIY+vJLXNDu5WTfiE9gq0Ze9aUACgiK50
5Q8KmRyBQ9D+9Xz9EPbSatI=
=GaMb
-----END PGP SIGNATURE-----