解析子流程Check_Output的表格输出

Question

I'm working on a Python(3.6) project in which I need to parse data from a tabular output of subprocess.check_output.

Here's How I'm using this: The output is:

NAME      HOSTS     ADDRESS           PORTS     AGE
mysvcs    *         107.178.250.150   80        1m

This out loads an empty column for ADDRESS like this:

NAME      HOSTS     ADDRESS           PORTS     AGE
mysvcs    *                           80        1m

But after few moments it loads the 107.178.250.150 in the 3rd column of second row. So, how can I get this address when it has been populated. Here's how I'm trying:

address = subprocess.check_output(["kubectl get ing | awk '{print $3}'"], shell=True)
            all_address = out_ip.decode().split('\n')
            print(all_address)
            address = all_address[1]
            print(address)
            if not address:
                while address == '':
                    out_address = subprocess.check_output(["kubectl get ing | awk '{print $3}'"], shell=True)
                    all_address = out_ip.decode().split('\n')
                    ip = all_address[1]
                    print(address)

But it's even not running the while loop and output as:

b'ADDRESS\\n80\\n' ['ADDRESS', '80', ''] 80

---

Update: Here's how I have tried stovfl's answer:

            ip = ''
            # Limit to 5 loops
            for n in range(5):
                result = subprocess.check_output(["kubectl get ing | awk '{print $3}'"], shell=True)
                # Make a List from the result, spliting at NewLine
                _list = result.decode().split('\n')

                # Check if Value 2 (Index==1) matches a IP Address
                if re.match(r'(\d+\.\d+\.\d+\.\d+)', _list[1]):
                    ip = _list[1]
                    print("match:{}".format(_list[1]))
                    # End the loop
                    break

                # Wait a second, before next try
                time.sleep(1)

                address = 'http://' + ip

Answer 1

Question : Parse tabular output, loop until results a IP Address

import time, re

# Limit to 5 loops
for n in range(5):
    result = subprocess.check_output(["kubectl get ing | awk '{print $3}'"], shell=True)
     # Make a List from the result, spliting at NewLine
    _list = result.decode().split('\n')
    # DEBUG
    print("list:{}".format(_list))

    # Check if Value 2 (Index==1) matches a IP Address
    if re.match(r'(\d+\.\d+\.\d+\.\d+)', _list[1]):
        print("match:{}".format(_list[1]))
        # End the loop
        break

    # Wait a second, before next try
    time.sleep(1)

Tested with Python: 3.4.2

Answer 2

If you are using Python 3.5 or newer, you can upgrade to subprocess.run() which has some useful features.

I would also factor out the shell and the Awk pipeline.

addr = None
while True:
    result = subprocess.run(
        ['kubectl', 'get', 'ing'],
        stdout=subprocess,PIPE, stderr=subprocess.PIPE,
        text=True, check=True)
    lines = result.stdout.split('\n')
    # [1:] - skip header line
    # [:-1] - skip empty element after last newline
    for line in lines[1:-1]:
        # replace Awk
        filelds = line.split()
        if len(fields) > 2 and '.' in fields[2]:
            addr = fields[2]
            break
    if addr:
        break
    time.sleep(1)

The text=True takes care of automatically normalizing the output to Unicode with standard line breaks. (Before Python 3.7 this was universal_newlines=True .)

Answer 3

Here's a working solution for Python(3.6):

addr = None
while addr is None:
     out = subprocess.check_output(["kubectl get ing | awk '{print $3}'"], shell=True)
     print(out)
     all_addr = out.decode().split('\n')
     if re.match(r"\d+\.\d+\.\d+\.\d+", str(all_addr[1])):
         addr = all_addr[1]
         print('Address is: {}'.format(addr))
     else:
         addr = None