Hi Thorsten,
One of our customers recently asked us about the different behavior of
constructs involving pipes in ksh scripts. I explained to them that
mksh spawns a sub-shell for the right hand side of each pipe, and
therefore whatever happens on that side of the pipe has no effect on
the main shell.
While I believe this is a perfectly sane implementation (and as a
matter of fact, bash and dash behave the same, at least by default), I
would like to make suggestions to the customer on how they can modify
their code so that it keeps working.
With bash, I would have a number of ideas. For example, they could set
option "lastpipe", which causes the last element of a pipe to be run in
the main shell, instead of the first. Or they could use process
substitution (the <() or >() syntax) to explicitly decide which part
goes in a sub-shell. But it doesn't seem mksh implements any of these?
One approach which works in all shells is to use temporary files
instead of pipes, however it could lower the performance of the
scripts, so I consider it a last resort solution.
Concretely, the customer's code looks like this:
command | while read line
do
if <condition>
then
exit
fi
process $line
done
... more stuff ...
The "exit" doesn't work in mksh as it ends the sub-shell, not the main
shell. One thing that would work in this case is exiting the sub-shell
with a specific error code, and checking for that code afterward:
command | while read line
do
if <condition>
then
ex
it 42
fi
process $line
done
if [ $? -eq 42 ]
then
exit
fi
... more stuff ...
I was wondering if there is any other trick you can suggest that would work in
mksh?
Thanks,
--
Jean Delvare
SUSE L3 Support
