Ah,? looks like this was already reported and discussed here a couple of
years ago:
https://stat.ethz.ch/pipermail/r-devel/2019-September/078478.html
Had no idea that "core of Bioconductor once took advantage of this
feature", seems that I've been fooled all these years ;-)
Sounds like this would deserve a big warning in ?`slot<-`
Thanks,
H.
On 22/03/2022 22:17, Herv? Pag?s wrote:> Hi,
>
> I just found out that `slot<-` performs an in-place modification of
> the object:
>
> ? setClass("A", slots=c(stuff="ANY"))
> ? x <- new("A", stuff=11:14)
> ? y <- `slot<-`(x, "stuff", value=99)
>
> Then:
>
> ? y
> ? # An object of class "A"
> ? # Slot "stuff":
> ? # [1] 99
>
> ? x
> ? # An object of class "A"
> ? # Slot "stuff":
> ? # [1] 99
>
> That doesn't seem right!
>
> Is this violation of the standard pass-by-value semantic somehow
> intended here because most of the time people are expected to do
> 'slot(x, "stuff") <- 99' instead, in which case the
violation cannot
> be observed so does not matter? I didn't find anything in the man page
> about this.
>
> I see this behavior with R 4.1.3 and current R devel.
>
> Thanks,
>
> H.
>
--
Herv? Pag?s
Bioconductor Core Team
hpages.on.github at gmail.com