R devel - Jan 2019 - setClass accepts slot-mismatch between slots and prototype arguments

I was installing the 'diffobj' package into TERR and got an error from
the
call
StyleSummary <- setClass("StyleSummary",
  slots=c(container="ANY", body="ANY", map="ANY"),
  prototype=list(
    container=function(x) sprintf("\n%s\n", paste0(x,
collapse="")),
    body=identity,
    detail=function(x) sprintf("\n%s\n", paste0("  ", x,
collapse="")),
    map=function(x) sprintf("\n%s", paste0("  ", x,
collapse="\n"))
  ))
because the prototype contained components not in the slots list.  R does
not complain about the mismatch, but new("StyleSummary") does name
make
something with a 'detail' slot.  Should this be an error?

I suspect that the package writer intended to include 'detail' in the
slots
argument.

Bill Dunlap
TIBCO Software
wdunlap tibco.com

	[[alternative HTML version deleted]]

Indeed that was on oversight on my part.? It is surprising that things like this
work:
> setClass('test', slots=c(a='ANY'), prototype=list(a=NULL,
b='hello'))
> new('test')@b
[1] "hello"
> slotNames(new('test'))
[1] "a"

I'm planning a release of diffobj right now so I will fix this, but I agree
that R should probably throw an error here.

Best,

B.




On Thursday, January 10, 2019, 11:30:13 AM EST, William Dunlap <wdunlap at
tibco.com> wrote:





I was installing the 'diffobj' package into TERR and got an error from
the call
StyleSummary <- setClass("StyleSummary",
? slots=c(container="ANY", body="ANY", map="ANY"),
? prototype=list(
? ? container=function(x) sprintf("\n%s\n", paste0(x,
collapse="")),
? ? body=identity,
? ? detail=function(x) sprintf("\n%s\n", paste0("? ", x,
collapse="")),
? ? map=function(x) sprintf("\n%s", paste0("? ", x,
collapse="\n"))
? ))
because the prototype contained components not in the slots list.? R does not
complain about the mismatch, but new("StyleSummary") does name make
something with a 'detail' slot.? Should this be an error?

I suspect that the package writer intended to include 'detail' in the
slots argument.

Bill Dunlap
TIBCO Software
wdunlap tibco.com

Thanks for the report. There is a comment from 2001 in the header for
reconcilePropertiesAndPrototype() that states:

"The prototype may imply slots not in the properties list.  It is not
required that the extends classes be define at this time.  Should it
be?"

But somewhere in the mid 2000's, I think, it became necessary for the
extends classes to be defined, so it may be safe to add this
constraint.

I'll commit a fix soon. While fixing, I found that the type
consistency check for slots shared with superclasses was completely
broken ever since it was written in 2002.

Michael

On Thu, Jan 10, 2019 at 8:31 AM William Dunlap via R-devel
<r-devel at r-project.org> wrote:
>
> I was installing the 'diffobj' package into TERR and got an error
from the
> call
> StyleSummary <- setClass("StyleSummary",
>   slots=c(container="ANY", body="ANY",
map="ANY"),
>   prototype=list(
>     container=function(x) sprintf("\n%s\n", paste0(x,
collapse="")),
>     body=identity,
>     detail=function(x) sprintf("\n%s\n", paste0("  ",
x, collapse="")),
>     map=function(x) sprintf("\n%s", paste0("  ", x,
collapse="\n"))
>   ))
> because the prototype contained components not in the slots list.  R does
> not complain about the mismatch, but new("StyleSummary") does
name make
> something with a 'detail' slot.  Should this be an error?
>
> I suspect that the package writer intended to include 'detail' in
the slots
> argument.
>
> Bill Dunlap
> TIBCO Software
> wdunlap tibco.com
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel