A lot of my scripts need a simple way to run shell commands. So I went ahead and wrote a method that makes it super easy to execute commands, and keep the output of both STDOUT and STDERR logged. Basically, just build a tuple of commands, such as:
args = ["ls", "-l"]
This builds a command to be executed, starting with the executable “ls” (which lists the contents of the current directory), and the argument “-l” (which lists extra information about the contents).
You can execute the script at the bottom like this:
python shell.py ls -l
And it will execute just like running the normal “ls -l” command, except it will redirect all output to two files in the current directory, named output.log and error.log.
import os, subprocess, sys
def main(argv):
run(argv)
def run(arg):
"""
Runs the given arg at the command line using the default shell. Outputs
when commands are run successfully.
Based on http://developer.spikesource.com/wiki/index.php/How_to_invoke_subprocesses_from_Python
@param Tuple args
A tuple of args, with the first being the command to be run, and
the remaining ones flags and arguments for the command. STDOUT and
STDERR are piped to tuple, waiting until the output is finished,
then writing both to the log files, if not empty.
Ex. ['apt-get', '-y', 'install', 'dnsmasq'], which installs
dnsmasq using apt-get, and assumes yes to questions.
"""
# Open output and write some info about the command to be written, including
# name of command and arguments.
# This could be modified to adjust how much is printed via a DEBUG variable.
with open(os.path.join(os.curdir, "output.log"), 'a') as outFile:
outFile.write("Command: ")
for a in arg:
outFile.write(a,)
outFile.write(" ")
outFile.write("\n")
# Open output and error log file and append to them the output of the commands
with open(os.path.join(os.curdir, "output.log"), 'a') as outFile:
with open(os.path.join(os.curdir, "error.log"), 'a') as errorFile:
# Call the subprocess using convenience method
retval = subprocess.call(arg, -1, None, None, outFile, errorFile)
# Check the process exit code, print error information if it exists
if not retval == 0:
errData = errorFile.read()
raise Exception("Error executing command: " + repr(errData))
if __name__=="__main__":
main(sys.argv[1:])
If you want to use it in your own scripts, just insert the run method, and pass the appropriate arguments to it. One potential change that might be helpful would be to move the log files to /var/log with the rest of the Linux logs. However normal users don’t usually have write permissions in /var/log, so an install wrapper would be helpful to both make the 2 files with writable permissions, and to move the file to /usr/bin so it can more easily be executed from the command line.
About Josh Gachnang
