[Rd] apropos changes in r-devel: intended feature or bug?

Martin Maechler maechler at stat.math.ethz.ch
Fri Dec 22 14:38:36 CET 2006


Ok, so be it:

We have seen that both apropos() and find()
have used `some' non-standard evaluation up to R 2.4.1
which gave quite incosistent behavior.

Getting rid of non-standard evaluation get's rid of all
inconsistencies but of course is not back-compatible either.

I'll do this.
Martin


>>>>> "Luke" == Luke Tierney <luke at stat.uiowa.edu>
>>>>>     on Fri, 22 Dec 2006 07:08:44 -0600 (CST) writes:

    Luke> On Fri, 22 Dec 2006, Kurt Hornik wrote:
    >>>>>>> Martin Maechler writes:
    >> 
    >>>>>>> "DeepS" == Deepayan Sarkar <deepayan.sarkar at gmail.com>
    >>>>>>> on Thu, 21 Dec 2006 22:07:27 -0800 writes:
    >> 
    DeepS> The old apropos started with: if
    DeepS> (!is.character(what)) what <-
    DeepS> as.character(substitute(what))
    >> 
    DeepS> The new one has:
    >> 
    DeepS> if (character.only) stopifnot(is.character(what))
    DeepS> else what <- as.character(substitute(what))
    >> 
    DeepS> i.e., the check for is.character(what) is
    DeepS> missing. This has the effect that 'what' can no
    DeepS> longer be a character string generated by a function
    DeepS> call unless 'character.only = TRUE'. I don't think
    DeepS> this was intended; the change makes previously valid
    DeepS> use invalid and I can't think of a situation where it
    DeepS> is useful.
    >> 
    >>> [ Did you read the corresponding NEWS entry? ]
    >> 
    >>> It now parallelizes the use in library() , require() etc,
    >>> and in particular does what the documentation says it does!
    >> 
    >>> The old behavior was much less consistent and not according to
    >>> documentation:
    >> 
    >>> apropos(lm)    was equivalent to apropos("lm")
    >>> but  apropos(fit)   gave an error.
    >> 
    >> I would actually prefer if we only had standard evaluation for apropos()
    >> and find().
    >> 
    >> (I understand we cannot do this for library() and help().)

    Luke> I agree completely.  If it is OK to make changes that make previous
    Luke> usage fail then it would be better to go to standard evaluation and
    Luke> let apropos(lm) fail.

    Luke> luke

    Luke> -- 
    Luke> Luke Tierney
    Luke> Chair, Statistics and Actuarial Science
    Luke> Ralph E. Wareham Professor of Mathematical Sciences
    Luke> University of Iowa                  Phone:             319-335-3386
    Luke> Department of Statistics and        Fax:               319-335-3017
    Luke> Actuarial Science
    Luke> 241 Schaeffer Hall                  email:      luke at stat.uiowa.edu
    Luke> Iowa City, IA 52242                 WWW:  http://www.stat.uiowa.edu

    Luke> ______________________________________________
    Luke> R-devel at r-project.org mailing list
    Luke> https://stat.ethz.ch/mailman/listinfo/r-devel



More information about the R-devel mailing list