问题
I want to lookup some data of running processes, actually I'm searching for the scratch directory to a corresponding job-ID. I managed to do that manually by the following commands (assuming the job-ID is 12345):
find ~ -name '12345.out'
This finds the output file according to the job:
/home/username/somefolder/12345.out
The scratch directory is written to a file named .scrdir in that folder, so
cd | cat /home/username/somefolder/.scrdir
brings me where I want.
I now want to combine that to a csh script or some alias in my .cshrc.
My atempt so far:
#/bin/csh
set jobfile = `find ~ -name $argv[1].out`
set jobdir = `dirname $jobfile`
set scrdir = `cat $jobdir/.scrdir`
echo $scrdir
where the first argument value is a job-ID, 12345 e.g. This script prints the right scratch directory but I want it to change my actual working directory to the scratch directory after I've called the script. How can I change the working directory from within a csh script?
I would also be grateful for any advice to refine the question / -title.
回答1:
Since there are no shell functions in csh, the only solution I found is to call the script with source.
The script:
#/bin/csh
set jobfile = `find ~ -name $argv[1].out`
set jobdir = `dirname $jobfile`
set scrdir = `cat $jobdir/.scrdir`
cd $scrdir
To call the script:
source myscript.sh
来源：https://stackoverflow.com/questions/28135356/change-working-directory-from-csh-script