Parsing a command output - Python

Question

I'm running a utility that parses the output of the df command. I capture the output and send it to my parser. Here's a sample:

Filesystem                512-blocks      Used  Available Capacity iused      ifree %iused  Mounted on
/dev/disk2                1996082176 430874208 1564695968    22% 2429281 4292537998    0%   /
devfs                            668       668          0   100%    1156          0  100%   /dev
map -hosts                         0         0          0   100%       0          0  100%   /net
map auto_home                      0         0          0   100%       0          0  100%   /home

Here's the function:

def parse_df(self, content):
    """Parse the `df` content output

    :param content: The command content output
    :return: (list) A list of objects of the type being parsed
    """
    entries = []
    if not content:
       return entries
    # Split the content by line and check if we should ignore first line
    for line in content.split("\n"):
        if line.startswith("Filesystem"):
            continue
        tokens = line.split()
        print tokens

However I'm getting the following output:

['/dev/disk2', '1996082176', '430876480', '1564693696', '22%', '2429288', '4292537991', '0%', '/']
['devfs', '668', '668', '0', '100%', '1156', '0', '100%', '/dev']
['map', '-hosts', '0', '0', '0', '100%', '0', '0', '100%', '/net']
['map', 'auto_home', '0', '0', '0', '100%', '0', '0', '100%', '/home']

The issue is map -host is supposed to be a single element (for the Filesystem column). I've tried to apply a regex tokens = re.split(r'\s{2,}', line) but the result was still not correct:

['/dev/disk2', '1996082176 430869352 1564700824', '22% 2429289 4292537990', '0%', '/']

What would be the correct way to parse the output?

You need to use a different delimiter maybe like `\t`? Even multiple spaces should work. — Nishant, Jan 09 '17 at 05:17
Each column has a fixed width. You could try splitting based on that — OneCricketeer, Jan 09 '17 at 05:18
@Nishant: Splitting by `\t`: `['/dev/disk2 1996082176 430874728 1564695448 22% 2429300 4292537979 0% /']` — cybertextron, Jan 09 '17 at 05:20
The filesystem is the first however many characters of a line... For example. http://stackoverflow.com/questions/4914008/how-to-efficiently-parse-fixed-width-files — OneCricketeer, Jan 09 '17 at 05:21
Sounds like a job for regular expressions; or [`os.statvfs`](https://docs.python.org/release/2.6.8/library/os.html#os.statvfs). — Burhan Khalid, Jan 09 '17 at 05:41
Unrelated, but there are system calls (e.g. statvfs) that will probably get what you want more directly. — Keith, Jan 09 '17 at 06:10

score 2 · Answer 1 · answered Jan 09 '17 at 05:57

Just split on one or more spaces which was followed by a digit or /

>>> import re
>>> s = '''/dev/disk2                1996082176 430874208 1564695968    22% 2429281 4292537998    0%   /
devfs                            668       668          0   100%    1156          0  100%   /dev
map -hosts                         0         0          0   100%       0          0  100%   /net
map auto_home                      0         0          0   100%       0          0  100%   /home'''.splitlines()
>>> for line in s:
    print re.split(r'\s+(?=[\d/])', line)


['/dev/disk2', '1996082176', '430874208', '1564695968', '22%', '2429281', '4292537998', '0%', '/']
['devfs', '668', '668', '0', '100%', '1156', '0', '100%', '/dev']
['map -hosts', '0', '0', '0', '100%', '0', '0', '100%', '/net']
['map auto_home', '0', '0', '0', '100%', '0', '0', '100%', '/home']
>>>

score 1 · Answer 2 · answered Jan 09 '17 at 05:33

If that is the behavior that you want, the easiest way I can see is to join the first element of the array until you reach a numeric element.

So something like this:

tokens = line.split()
n = 1
while n < len(tokens) and not tokens[n].isdigit():
    n += 1
tokens[0] = ' '.join(tokens[:n])
tokens = [ tokens[0] ] + tokens[n:]

Alternatively you could try @cricket_007’s suggestion:

first_token = line[:15].strip()
tokens = [ first_token ] + [ x.strip() for x in line[15:].split() ]

Nishant · Answer 3 · 2017-01-09T09:38:05.173

Since FS is going to probably have multiple spaces and as long as you can pre-determine that you can split using different delimiters and combine them eventually.

fs, rest = re.split(r'\s{2,}', line, 1)
result = [fs] + rest.split()

But this won't work is fs is separated by a single space like a big one.

Agree with comments that using os.statvfs(path) is a better tool for this. df would be a subprocess call.

Parsing a command output - Python

3 Answers3