[R] Recovering object names when using the ... argument in a fn XXXX

[Duncan Murdoch]

On 13-10-18 1:54 PM, Bert Gunter wrote:
> Yes, similar, but better, as match.call() will get unwanted named
> arguments, too.
>
> However, I do not understand the
>
> substitute(...())
>
> idiom. Would you care to explain it? (No is an acceptable  answer!).

I suspect it's a bug, though I can't see it's one that's likely to need 
fixing.  The general idea is that a function call like ...() is parsed 
into a pairlist containing just the name "...", then substitute replaces 
it with the content of that variable, which is a pairlist containing the 
unevaluated argument list.  So by that analysis, you might expect to get 
the same result using

pairlist(...)

However, you don't, because the latter expression evaluates all the 
arguments to the function, while Bill's idiom leaves them unevaluated. 
I can't think of any documented reason why that should be, but on the 
other hand, I can't think of any reason it would cause problems.  So I'd 
say it's unlikely to be deliberately changed, but it might change as a 
result of some internal change to R.

Duncan Murdoch


>
> I would have expressed it as:
>
> as.list(substitute(list(...)))[-1]
>
> to convert the parse tree to a list. (which is again better than using
> match.call() ).
>
> Best,
> Bert
>
> On Fri, Oct 18, 2013 at 10:27 AM, William Dunlap <wdunlap at tibco.com> wrote:
>>> I am using the ... argument to parmeterize a user define fn to accept
>>> multiple input objects. I subsquently save all these data as a list.
>>> Question: what is the best way to recover or extract the original object
>>> names that were fed to the fn?
>>
>> The following function, ellipsisInfo, returns character strings representing the
>> actual arguments to the function.  If the function was called with tags on the
>> arguments, as in ellipsisInfo(tag=argument), it makes those tags the names
>> on the returned character  vector.  It does not evaluate the ... arguments, so
>> you don't run into problems with evaluating arguments too soon or evaluating
>> ones that should not be evaluated most of the time.
>>
>> ellipsisInfo <- function(...) {
>>      # get the unevaluated expressions given as arguments
>>      unevaluatedArgs <- substitute(...())
>>      # convert those expressions to text (truncate to single line)
>>      unevaluatedArgsAsText <- vapply(unevaluatedArgs, function(a)deparse(a)[1], "")
>>      unevaluatedArgsAsText
>> }
>>
>> E.g.,
>>
>>> i <- ellipsisInfo(x, log(10), e=exp(1), onProblem=stop("there was a problem"))
>>> i
>>
>>                              "x"
>>
>>                        "log(10)"
>>                                e
>>                         "exp(1)"
>>                        onProblem
>> "stop(\"there was a problem\")"
>>> ifelse(names(i)=="", i, names(i)) # use tag if supplied, otherwise argument itself
>> [1] "x"         "log(10)"   "e"
>> [4] "onProblem"
>>
>> Bill Dunlap
>> Spotfire, TIBCO Software
>> wdunlap tibco.com
>>
>>
>>> -----Original Message-----
>>> From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf
>>> Of Dan Abner
>>> Sent: Friday, October 18, 2013 9:06 AM
>>> To: r-help at r-project.org
>>> Subject: [R] Recovering object names when using the ... argument in a fn XXXX
>>>
>>> Hi all,
>>>
>>> I am using the ... argument to parmeterize a user define fn to accept
>>> multiple input objects. I subsquently save all these data as a list.
>>> Question: what is the best way to recover or extract the original object
>>> names that were fed to the fn?
>>>
>>> Thanks,
>>>
>>> Dan
>>>
>>>        [[alternative HTML version deleted]]
>>>
>>> ______________________________________________
>>> R-help at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-help
>>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>>> and provide commented, minimal, self-contained, reproducible code.
>>
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>
>
>