Question

I am working on this networking project for which I need to create many shell scripts to run tools like “tc”, “arp” and “ping” from multiple nodes to multiple nodes at a time. I need to generate several such scripts for many nodes. So I need to create many scripts for each scenario and configuration, each of which varies slightly from the other. Also, my work also deals with Mininet. I need to create python scripts to generate many custom topologies there. It is a pain to hard code so many long scripts.

So I wanted to know what is the best way to automate the creation of these scripts? Is it possible to generate these shell/python scripts using another shell script? Or should I generate these using Java? Which is the easiest approach? Just want to your opinions before I start working. Thanks 🙂

Answer 1

Here’s a python function that creates a script used to set to a hyperlink in Excel. The new script opens a particular pdf at a particular page. The larger program to which this function belongs generates dozens, sometimes hundreds, of these scripts per run, all of which are identical except for the args passed to this function. They are all automatically deleted when the user closes Excel and exits the program:

def makeOpener(pdf, page):
    '''makes a pyw file to use to hyperlink in excel'''
    t = 'c:/temp'
    if not os.path.isdir(t):
        os.makedirs(t)
    tempfile.tempdir = t
    py = tempfile.NamedTemporaryFile(suffix='.pyw', delete=False)

    lst = ['import subprocess, win32api',
           'pdf = r"' + pdf + '"',
           'page = "' + str(int(page)) + '"',
           '_, exe = win32api.FindExecutable(pdf)', 
           'opts = "page=" + page', 
           'clst = [exe, "/A", opts, "/N", pdf]', 
           'DP = 0x00000008', 
           'subprocess.Popen(clst, shell=False, creationflags=DP)'
           ]
    py.write('\n'.join(lst))       
    py.close()
    return py.name

Mike