Opening a terminal application from python and running custom scripts inside it

Question

I'm working with a software called dc_shell that has a terminal command (also called dc_shell) on a CentOS Linux server. when I run dc_shell command, I'm connected to its terminal and I'm able to run scripts/commands inside it. (This is all done manually)

So the real problem is that I want to do this task all from a Python program. Meaning that I have a Python code which does some task, and after that has to open dc_shell and run some commands inside it.

I have used subprocess.Popen before and this doesn't have any problem when I run commands like ls or other general terminal commands. But when I run dc_shell command it seems like it crashes and nothing happens, and when I try to terminate the session I get the following errors in my terminal.

Here's my code:

def run_scripts():
    commandtext = 'cd ..; dc_shell-xg-t; set_app_var link_library "slow.db"; set_app_var target_library "slow.db"; set_app_var symbol_library "tsmc18.sdb";'
    print(commandtext)
    process = subprocess.Popen(commandtext,stdout=subprocess.PIPE, shell=True)
    proc_stdout = process.communicate()[0].strip()
    print(proc_stdout)

and the output is:

cd ..; dc_shell-xg-t; set_app_var link_library "slow.db"; set_app_var target_library "slow.db"; set_app_var symbol_library "tsmc18.sdb";

and nothing happens... and after terminating I get:

[User@server python]$ /bin/sh: set_app_var: command not found
/bin/sh: set_app_var: command not found
/bin/sh: set_app_var: command not found

Answer 1

Do you need to use dc_shell to run your commands?

If so, that should be your executable and the rest of commands your arguments.

You should never use shell=True due to security considerations (the warning in the 2.x docs for subprocess seems much clearer to me).