R devel - Jun 2005 - Bug in new() or validObject() in methods package (PR#7922)

Something in new() or validObject() in the methods package is messing 
up.  This happens in both 2.1.0 and R-devel in Windows.


I'd like to have an empty expression for a slot in a class.  An empty
expression is an expression:

 > is.expression(expression())
[1] TRUE
 >
 > is(expression(), "expression")
[1] TRUE
 > class(expression())
[1] "expression"


However, when I try to use this as the initial value, I get an error
thrown back at me:

 > setClass("foo", representation(bar = "expression"))
[1] "foo"
 >
 > new("foo", bar = expression())
Error in validObject(.Object) : invalid class "foo" object: invalid 
object for slot "bar" in class "foo": got class
"NULL", should be or
extend class "expression"

Is it doing an extra eval or something?  It looks like it:

 > x <- expression()
 > x
expression()
 > eval(x)
NULL
 >
 > x <- quote(expression())
 > x
expression()
 > eval(x)
expression()
 > eval(eval(x))
NULL

However, it doesn't always happen:

 > new("foo", bar = expression(1+1))
An object of class "foo"
Slot "bar":
expression(1 + 1)

 > x <- expression(1+1)
 > x
expression(1 + 1)
 > eval(x)
[1] 2

Duncan Murdoch

murdoch@stats.uwo.ca writes:
> Is it doing an extra eval or something?  It looks like it:
> 
>  > x <- expression()
>  > x
> expression()
>  > eval(x)
> NULL
>  >
>  > x <- quote(expression())
>  > x
> expression()
>  > eval(x)
> expression()
>  > eval(eval(x))
> NULL
You do realize that the two expression() results are not identical:
> x <- quote(expression())
> class(x)
[1] "call"
> x <- expression()
> class(x)
[1] "expression"

Not that I can fathom what bearing that has on the real problem...

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907

The bug might be here:
> is.null(expression())
[1] TRUE

But
> is.null(expression(NULL))
[1] FALSE

So it might look to the methods package like you're passing in a NULL
value for @bar. I might argue that expression() should not be NULL (and
only NULL is NULL) as I have had similar trouble trying to extend
language objects to S4 classes, partially because length 0 name, call
and expression objects are either not available, or in this case,
handled poorly.

Robert

-----Original Message-----
From: Duncan Murdoch [mailto:murdoch@stats.uwo.ca] 
Sent: Monday, June 06, 2005 4:58 PM
To: Peter Dalgaard
Cc: r-devel@stat.math.ethz.ch
Subject: Re: [Rd] Bug in new() or validObject() in methods package
(PR#7922)


On 6/6/2005 4:55 PM, Peter Dalgaard wrote:
> murdoch@stats.uwo.ca writes:
> 
>> Is it doing an extra eval or something?  It looks like it:
>> 
>>  > x <- expression()
>>  > x
>> expression()
>>  > eval(x)
>> NULL
>>  >
>>  > x <- quote(expression())
>>  > x
>> expression()
>>  > eval(x)
>> expression()
>>  > eval(eval(x))
>> NULL
> 
> You do realize that the two expression() results are not identical:
> 
>> x <- quote(expression())
>> class(x)
> [1] "call"
>> x <- expression()
>> class(x)
> [1] "expression"
> 
> Not that I can fathom what bearing that has on the real problem...
I figured they weren't identical, given that they gave different results

  when eval'd, but I didn't know what the difference was exactly.
Thanks!

Duncan

______________________________________________
R-devel@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel