[R] Object validation and formal classes

[Torsten Steuernagel]

On 30 Jan 2004 at 14:05, John Chambers wrote:

> It was never the intention that validity checking happen automatically
> on _every_ assignment of an object from the class--since those
> assignments take place frequently during evaluation of functions, the
> overhead would be unacceptable.  And as Robert points out, one needs
> to postpone validity checking until a set of mutually dependent
> changes is finished.

Thank's for this clarification. I'll try to find another solution. Could you 
comment on the "access" argument in "setClass()" ? From the 
"setClass()" reference I can only guess what that might be:

"access: Access list for the class.  Saved in the definition, but not
          currently used."

Is this intended to implement an access control mechanism similar to 
C++ like I stated in my previous message ? If that's the case I'll simply 
provide access/assign methods for the slots that need validation. Once 
access control is available I could hide those slots.

> The "Programming with Data" description says that validity checking
> takes place on "permanent assignment".  But this was not written with
> R in mind, and the idea of permanent assignment is ambiguous in R.  It
> could mean all assignments into the Global environment, or it could
> mean on serializing (saving the workspace, e.g.).  If we agree on a
> useful interpretation, automatic validity checking might be reasonable
> in that sense in the future.

I must admit that I didn't look into "Programming with Data", which I 
don't have available right now, but I believe "S Programming" says the 
same. I think the example code in the reference for "setValidity()" in R 
is misleading here, especially the last four lines:

## Now we do something bad
t1 at x <- 1:20
## This should generate an error
## Not run: try(validObject(t1))

- Torsten