[Rd] Accessing the formals() of a non-exported method without :::?

[Gavin Simpson]

By

"...I had forgotten that a package can refer to any of *its own packages*"

I of course meant

"...I had forgotten that a package can refer to any of *its own functions*"

Guess I should call it a night...

G

On 23 August 2013 21:49, Gavin Simpson <ucfagls at gmail.com> wrote:
> Thanks Michael, somewhat embarrassingly, I had forgotten that a
> package can refer to any of *its own packages*. So I don't need
> analogue:::foo.bar as that is being called from a function in the
> analogue namespace.
>
> Cheers,
>
> G
>
> On 23 August 2013 09:27, R. Michael Weylandt
> <michael.weylandt at gmail.com> <michael.weylandt at gmail.com> wrote:
>>
>>
>> On Aug 23, 2013, at 11:15, Gavin Simpson <ucfagls at gmail.com> wrote:
>>
>>> Dear List,
>>>
>>> I'm in the process of making tweaks to my various R packages following
>>> changes in r-devel for package checks. I'm wondering about the one use
>>> of ::: in one of my packages. I am arranging for a call to a
>>> non-exported S3 method via do.call. For this I need the arguments of
>>> the function and hence I was doing
>>>
>>>    Args <- head(formals(analogue:::wa.default), -1)
>>>
>>> Following the recent thread on legitimate uses of ::: I think the
>>> above is both acceptable and won't generate a Note now with R CMD
>>> check following a recent change to that code. But is there a better
>>> way to get the formal arguments of a non-exported S3 method?
>>>
>>
>> Something like (untested)
>>
>> formalsS3 <- function(...) formals(getS3method(...))
>>
>> might work. I haven't checked to see how R CMD check feels about getS3method() though.
>>
>> Michael
>>
>>
>>> Thanks.
>>>
>>> G
>>>
>>> --
>>> Gavin Simpson
>>>
>>> ______________________________________________
>>> R-devel at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>
>
> --
> Gavin Simpson, PhD



-- 
Gavin Simpson, PhD