Correct way to check if command executed successfully

[Cecil Westerhof]

After being disconnected from the internet for six hours I decided to
write something to check my internet connection. The core would be:
if {[catch {exec ping -q -c 1 -W 1 ${pingDomain} >&/dev/null}]} {
puts "NOT connected"
} else {
puts Connected
}

Is this the best way to check if a command executed successfully? That
is the only thing I am interested in: was it executed successfully or
not.

Or maybe I should change the first line to:
set cmd [list exec ping -q -c 1 -W 1 ${pingDomain} >&/dev/null]
if {[catch ${cmd}]} {

--
Cecil Westerhof
Senior Software Engineer
LinkedIn: http://www.linkedin.com/in/cecilwesterhof

[stefan]

> Is this the best way to check if a command executed successfully?

set fh [open |[list ping -c 1 example.org]]
if {[catch {close $fh} e opts]} {
lassign [dict get $opts -errorcode] msg pid exitcode
puts $msg
puts $pid
puts $exitcode
}

open/close will give you access to the child process' details. You can then watch out for CHILDSTATUS, CHILDKILLED & friends.

On a different page: When the aim is to test internet connectivity, why not use Tcl's [socket]. Btw., what you are after has been done before, check the wiki: http://wiki.tcl.tk/41924

HTH,
Stefan

[Cecil Westerhof]

stefan <stefan....@wu.ac.at> writes:

>> Is this the best way to check if a command executed successfully?
>
> set fh [open |[list ping -c 1 example.org]]
> if {[catch {close $fh} e opts]} {
> lassign [dict get $opts -errorcode] msg pid exitcode
> puts $msg
> puts $pid
> puts $exitcode
> }
>
> open/close will give you access to the child process' details. You can
> then watch out for CHILDSTATUS, CHILDKILLED & friends.

Looks interesting. If I do it in the shell it seems to work. But if I
do it in a program I get:
CHILDKILLED
12516
SIGPIPE

> On a different page: When the aim is to test internet connectivity, why
> not use Tcl's [socket]. Btw., what you are after has been done before,
> check the wiki: http://wiki.tcl.tk/41924

I will look into those.

[Cecil Westerhof]

stefan <stefan....@wu.ac.at> writes:

>> Is this the best way to check if a command executed successfully?
>
> set fh [open |[list ping -c 1 example.org]]
> if {[catch {close $fh} e opts]} {
> lassign [dict get $opts -errorcode] msg pid exitcode
> puts $msg
> puts $pid
> puts $exitcode
> }
>
> open/close will give you access to the child process' details. You can
> then watch out for CHILDSTATUS, CHILDKILLED & friends.

That does not seem to work. (See my other reply.) But I used to better
some things.

I now initialise the following:
set cmd [list exec ping -c 1 -W 1 ${pingDomain} >/dev/null]
set intervalCheck 1
set notConnected "ping: ${pingDomain}: Temporary failure in name resolution"

And have the following proc defined:
proc checkInternetConnection {} {
if {[catch ${::cmd} errorStr opts]} {
if {${errorStr} == ${::notConnected}} {
set message "NOT connected"
} else {
set message "OTHER ERROR: ${errorStr}"
}
} else {
set message Connected
}
showMessage ${message} True
after [getWaitMinutesTicks ${::intervalCheck}] checkInternetConnection
}

showMessage is to show a message. For example:
13:35:00: Connected
13:36:40: NOT connected
13:37:40: NOT connected
13:38:00: Connected

And I use:
after [getWaitMinutesTicks ${::intervalCheck}] checkInternetConnection

To make sure that the check is run at the start of every minute.