R devel - Jan 2013 - Bug in list subset assignment due to NAMED optimization

In R version 2.15.2 (2012-10-26) i386-apple-darwin9.8.0/i386 (32-bit) I get
the following:
> a <- list(1)
> (a[[1]] <- a)
[[1]]
[[1]][[1]]
[1] 1

but
> a <- list(1)
> b <- a
> (a[[1]] <- a)
[[1]]
[1] 1

And similarly:
> a <- list(x=1)
> (a$x <- a)
$x
$x$x
[1] 1

but
> a <- list(x=1)
> b <- a
> (a$x <- a)
$x
[1] 1

In both cases the result of the first sequence is wrong. It's returning the
updated `a` rather than the RHS of the assignment. The second sequence in
both cases is correct; the assignment to `b` increments the NAMED value
causing the necessary copy to be made so the RHS is returned from the
assignment.

Would it be sufficient to add a check to do_subassign2_dflt
and do_subassign3_dflt that creates a duplicate of the LHS if the LHS & RHS
are the same object?

Justin

	[[alternative HTML version deleted]]

On Tue, 8 Jan 2013, Justin Talbot wrote:
> In R version 2.15.2 (2012-10-26) i386-apple-darwin9.8.0/i386 (32-bit) I get
> the following:
>
>> a <- list(1)
>> (a[[1]] <- a)
> [[1]]
> [[1]][[1]]
> [1] 1
>
> but
>
>> a <- list(1)
>> b <- a
>> (a[[1]] <- a)
> [[1]]
> [1] 1
>
> And similarly:
>
>> a <- list(x=1)
>> (a$x <- a)
> $x
> $x$x
> [1] 1
>
> but
>
>> a <- list(x=1)
>> b <- a
>> (a$x <- a)
> $x
> [1] 1
>
> In both cases the result of the first sequence is wrong. It's returning
the
> updated `a` rather than the RHS of the assignment. The second sequence in
> both cases is correct; the assignment to `b` increments the NAMED value
> causing the necessary copy to be made so the RHS is returned from the
> assignment.
>
> Would it be sufficient to add a check to do_subassign2_dflt
> and do_subassign3_dflt that creates a duplicate of the LHS if the LHS &
RHS
> are the same object?
Thanks -- that's probably the best fix but I want to think it thorugh
a bit more before committing.

Best,

luke
>
> Justin
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
-- 
Luke Tierney
Chair, Statistics and Actuarial Science
Ralph E. Wareham Professor of Mathematical Sciences
University of Iowa                  Phone:             319-335-3386
Department of Statistics and        Fax:               319-335-3017
    Actuarial Science
241 Schaeffer Hall                  email:   luke-tierney at uiowa.edu
Iowa City, IA 52242                 WWW:  http://www.stat.uiowa.edu

Thanks for the report.  This is now fixed in R-devel and R-patched
by increasing NAMED on the RHS value to 2 when it is not 0.

Best,

luke

On Tue, 8 Jan 2013, Justin Talbot wrote:
> In R version 2.15.2 (2012-10-26) i386-apple-darwin9.8.0/i386 (32-bit) I get
> the following:
>
>> a <- list(1)
>> (a[[1]] <- a)
> [[1]]
> [[1]][[1]]
> [1] 1
>
> but
>
>> a <- list(1)
>> b <- a
>> (a[[1]] <- a)
> [[1]]
> [1] 1
>
> And similarly:
>
>> a <- list(x=1)
>> (a$x <- a)
> $x
> $x$x
> [1] 1
>
> but
>
>> a <- list(x=1)
>> b <- a
>> (a$x <- a)
> $x
> [1] 1
>
> In both cases the result of the first sequence is wrong. It's returning
the
> updated `a` rather than the RHS of the assignment. The second sequence in
> both cases is correct; the assignment to `b` increments the NAMED value
> causing the necessary copy to be made so the RHS is returned from the
> assignment.
>
> Would it be sufficient to add a check to do_subassign2_dflt
> and do_subassign3_dflt that creates a duplicate of the LHS if the LHS &
RHS
> are the same object?
>
> Justin
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
-- 
Luke Tierney
Chair, Statistics and Actuarial Science
Ralph E. Wareham Professor of Mathematical Sciences
University of Iowa                  Phone:             319-335-3386
Department of Statistics and        Fax:               319-335-3017
    Actuarial Science
241 Schaeffer Hall                  email:   luke-tierney at uiowa.edu
Iowa City, IA 52242                 WWW:  http://www.stat.uiowa.edu