Skip to content

"CTRL-C" and "try"

2 messages · lutz.thieme@amd.com, Luke Tierney

Original
lutz.thieme@amd.com
# 
Hello everybody,

if I'm running a R-script on a command line R-session which uses the function "try" and I'd like
to interrupt the execution of my R-script pressing "CTRL-C" than R ignores the first "CTRL-C"
key stroke and after I pressed "CTRL-C" second time R by itself is interrupted and I fall back 
to the (UNIX-) command prompt. Without using "try" all works fine and I'm able to interrupt my
R-script. How can I avoid this behavior?

platform sparc-sun-solaris2.8
arch     sparc               
os       solaris2.8          
system   sparc, solaris2.8   
status                       
major    1                   
minor    3.0                 
year     2001                
month    06                  
day      22                  
language R  

Thanks in advance - regards,

Lutz

Lutz Thieme
AMD Saxony Manfacturing GmbH
Product Engineering
phone:	+49 351 277-4269
fax:	+49 351 277-9-4269
email:	lutz.thieme at amd.com


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Luke Tierney
# 
The first one is probably not being ignored but is being caught by the
`try'.  This is the way `try' and C-c currently interact--it is not
clear this is the right thing to do or that this will remain true.
One of the things in the works is a more structured exception handling
system that will allow the interaction of try, C-c, and other
non-local exits to be specified more clearly.

I suspect you are using C-c again before R has a chance to get back to
the top level command loop.  I believe solaris is one of those systems
where signal handlers need to be re-installed after a signal is
received.  We currently do this at the top level loop but not at the
point where a `try' intercepts a jump out of a C-c.  As a result the
default signal handler is in place for the period between the first
C-c and return to the top level loop, and the default handler exits
when a C-c is received.  [I'm speculating a bit here since I don't
have easy access to a solaris system, but you should be able to
confirm it by running

{ try(repeat{}); cat("got through try\n"); repeat{}}

at the command line.  The first C-c should take you out of the `try'.
On my Linux system a second C-c takes me to the command line but on
your it probably exits R.]

luke
On Thu, Apr 11, 2002 at 03:17:30PM +0200, lutz.thieme at amd.com wrote: