RE: "If F is a partially applied function ..."

The implementation of F may well be an external function or any other kind of function.  The other bullet points say what it means to evaluate each kind of function.  What else would you like this to say?

Instead of "supplying the value of each fixed position as the argument of the corresponding parameter", I think it should say "supplying the value of each non-local variable binding as the argument of the corresponding parameter."  Would that fix the problem?  I agree that it's wrong as written.

Jonathan

________________________________________
From: Michael Dyck [jmdyck@ibiblio.org]
Sent: Tuesday, February 9, 2016 2:24 AM
To: public-xsl-query@w3.org
Subject: "If F is a partially applied function ..."

The proposed wording in 3.1.5.1 says:
      If F is a partially applied function, the implementation of F is
      called, supplying the value of each fixed position as the argument
      of the corresponding parameter.

"the implementation of F is called":
This phrase raises the question of what it means to call the implementation
of a function. Of course, if the implementation is implementation-dependent,
we can't say. But if it isn't, we can, and should, but I don't believe this
text does so.

"supplying the value of each fixed position as the argument of the
corresponding parameter."
First, fixed positions are a property of a partial function application
(i.e., a hunk of syntax), not of a partially applied function. When it comes
time to invoke the PAF, fixed positions don't exist.

Second, supplying a value for each fixed position is exactly what the PFA
did when it created the PAF. When you invoke the PAF, you instead supply
values for its parameters.

-Michael

Received on Tuesday, 9 February 2016 15:48:49 UTC