[R] connections: remember to close it!

[Prof Brian Ripley]

On Tue, 2 Oct 2001, Liaw, Andy wrote:

> Hi all,
>
> I have newly discovered the power of connections (although just a small part
> of it, namely the file() connection).  It allowed me to process data file
> with arbitrary number of fields per line, which I was using awk for
> preprocessing before readling into R.  The combination of the connection,
> readLines() and strsplit() really make the work a lot easier and more
> organized.  Thanks a bundle to the R Core for this excellent work!!
>
> One reminder for people new to connections:  If you open connections inside
> a function, make sure they are closed before exiting from that function,
> otherwise you end up with connections that can not be closed (or at least I
> tried closeAllConnections() without success on WinNT4).  Not being very
> smart or clear-headed, I turned a working R script that opens the connection
> and process the data into an R function, but introduced some bugs in the
> process.  I forgot to add "on.exit(close(filecon))" right after I open the
> connection in the function.  As a result, repeated failed calls to the
> function generated a bunch of open connections that I can't close.  I'm not
> sure what the consequences of this is.  I was able to quit R gracefully, but
> my NT seemed to have become unstable afterward.

Some terminology:

     `close' closes and destroys a connection.

     `closeAllConnections' closes all open user connections.

Notice the distinction.

You can close and destroy connections later. Just call
showConnections(all=FALSE)  and then use
close(getConnection(some.number)).  This *is* on the *same* help page as
closeAllConnections.

closeAllConnections was broken when close was made generic. It should read

function ()
{
    set <- getAllConnections()
    set <- set[set > 2]
    for (i in seq(along = set)) close(getConnection(set[i]))
    invisible()
}

and needs to be amended to handle the sink stack.

-- 
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 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._