I'm a beginner in Python, and I've been trying to call a command line app, but it fails:
>>> import subprocess as s
>>> s.call("gpio -g read 17")
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "/usr/lib/python2.6/subprocess.py", line 470, in call
return Popen(*popenargs, **kwargs).wait()
File "/usr/lib/python2.6/subprocess.py", line 623, in __init__
errread, errwrite)
File "/usr/lib/python2.6/subprocess.py", line 1141, in _execute_child
raise child_exception
OSError: [Errno 2] No such file or directory
But then if I add shell=True
it all starts working. Can someone explain why?
>>> import subprocess as s
>>> s.call("gpio -g read 17", shell=True)
>>> 0
You're not using call right. Look at the introduction or any of the examples in the docs. The first argument of call is "args", a sequence of arguments, where arg[0] is the program to run.
So, when you do this:
s.call("gpio -g read 17")
There are two ways subprocess could interpret this. It should run a program called "g" with arguments "p", "i", "o", " ", etc. (Remember, strings are sequences of characters.) It might instead run a program called "gpio -g read 17" with no additional arguments. Either way, it's not going to find such a program. (Unless you happen to have a program called "g" or "gpio -g read 17" on your PATH, in which case it'll do the wrong thing instead of giving you an error…)
What you want is:
s.call(["gpio", "-g", "read", "17"])
So, why does this work if you pass shell=True
? Because this whole string gets passed to the shell, which then does its own parsing of the command line and separates things by spaces. It's like calling os.system("gpio -g read 17")
.
Please note that all of the above is a bit oversimplified (it ignores Windows, and shell parsing isn't really just "separate by spaces", and so on), so you should actually read the documentation. (Also, whoever wrote the subprocess
docs is a better writer than me.)