RE: [MO-java-dev] ActiveCall class

[Tim M. <tim...@po...>]

The way I see it - you are (or want to) construct a MethodMatcher based on a
name or name constraint and parameters or parameter contraints (in dot net I
believe you additional get a return type as well).

I would expect this object to compare itself to a java Method object (at
invoke time) and if it matches, do something (e.g. throw an exception or
answer a value etc.) - the doing could involve a series of decorations as we
have now (although my hunch is that we can simplify these down a bit).

So the code that I see now - doesn't really work in this manner?

I also feel - given some of the questions/features that other people have
asked for, that matching is sometimes something people want different
policies for on a method by method basis (e.g. match once only, match one or
more) - (which combined with your excellent auto-return type stub, can give
useful behavior especially in a name match constraint scenario like
StartsWith("get"))

So I think I would take the pieces that we have and combine them in a much
different way.

Tim

> -----Original Message-----
> From: moc...@li...
> [mailto:moc...@li...]On Behalf Of
> Nat Pryce
> Sent: 05 July 2003 22:55
> To: Tim Mackinnon; Steve Freeman; 'Mockobjects-Java-Dev'
> Subject: Re: [MO-java-dev] ActiveCall class
>
>
> The name was meant to indicate that it represents the call currently
> "active" in the mock.  I agree that there is a better name.  Personally I
> would like to name it "Call" because all the other uses of the word "call"
> don't actually refer to calls but to things that can be called.
>
> Perhaps Invocation would be a better name.
>
> There is actually very little behaviour required, as far as I
> have used it.
> It doesn't need to be compared to a method: it is constructed *from* a
> method.  Other objects (matchers, stubs, etc.) examine the state that it
> carries.  I guess some of that behaviour could be moved into the
> ActiveCall
> class but that feels wrong to me.  Why should an invocation have knowledge
> of the behaviour caused by that invocation?
>
> Cheers,
>             Nat.
> _______________________
> Dr. Nathaniel Pryce
> B13media Ltd.
> http://www.b13media.com
> +44 (0)7712 526 661
>
> ----- Original Message -----
> From: "Tim Mackinnon" <tim...@po...>
> To: "Steve Freeman" <st...@m3...>; "'Mockobjects-Java-Dev'"
> <moc...@li...>
> Sent: Saturday, July 05, 2003 8:28 PM
> Subject: RE: [MO-java-dev] ActiveCall class
>
>
> > The idea of ActiveCall is a good one (I'm not too keen on the name
> myself -
> > why is it active?) - Actually more to the point I'm kind of suprised its
> got
> > no behavior at all (making it even less active?).
> >
> > I would expect it to do work in deciding whether it was equal
> to a Method
> > (e.g. what invoke() gets) ...???
> >
> > Tim

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.495 / Virus Database: 294 - Release Date: 30/06/2003