[Rd] R CMD versus Rcmd

Prof Brian Ripley ripley at stats.ox.ac.uk
Sun Feb 22 09:19:21 MET 2004


On Sat, 21 Feb 2004, Chris Lawrence wrote:

> Ross Ihaka wrote:
> 
> > Dirk Eddelbuettel wrote:
> >
> >> On Sat, Feb 21, 2004 at 01:22:41PM -0500, Duncan Murdoch wrote:
> >>
> >>> I've just committed changes to R-devel (to become 1.9.0) to allow
> >>> scripts to be invoked using the Unix-style "R CMD command" in Windows.
> >>> "Rcmd command" will still work, but is no longer documented.
> >>
> >
> > [ I don't seem to have the original handy ]
> >
> > This seems to have broken builds under linux (at least on Fedora).
> > I get a string of messages of the form
> >
> > /home/ihaka/R/bin/R CMD: /home/ihaka/R/bin/R CMD: No such file or 
> > directory
> >
> > I am currently working around this by changing lines in
> > src/scripts/R.sh from
> >
> >       exec sh "${R_HOME}/bin/R CMD" "${@}" ;;
> >
> > to
> >
> >       exec sh "${R_HOME}/bin/Rcmd" "${@}" ;;
> >
> > between "configure" and "make".
> 
> 
> Probably the ideal solution is:
> 
> exec sh "${R_HOME}/bin/R" CMD "${@}" ;;
> 
> Otherwise, sh tries to interpret the first argument as a pathname (i.e. 
> looking for "R CMD" in that directory, rather than treating the space as 
> an argument separator).

There was no problem with version 1.68, and I believe this was an error
(as otherwise the call would be infinitely recursive, R CMD calling R 
CMD).

I have reverted the change (in R.sh.in) to 1.68.

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595



More information about the R-devel mailing list