在defaultdict（list）python中计数

Question

I am not able to get it working without your help. I want to filter some syslog messages based on devicename. The output should look like this.

Device1: 1x failure1,50 x failure2, 20x failure3
Device3: 10 x failure1,5 x failure2, 2x failure3

Code:

frequencies = defaultdict(list)

word = ['syslog1error1','syslog1error2','syslog1error3']

def findpattern():
    for line in syslog:
            if re.search(r"regexforhostname",line):
                hostname= line.strip()[16:27]
                for failure in word:
                    if failure in line:     
                    frequencies[hostname].append(failure)

x = findpattern()

print frequencies

Output looks like

'Devicename':'syslog1error1', 'syslog1error1', 'syslog1error2', 'syslog1error3'

I would like to count the double entries in the list. But I can't get it running with import collections (counter)

Please help.

Answer 1

Use collections.Counter() (see Counter in Collections module Python if you are on a Python version < 2.7):

from collections import Counter, defaultdict

def findpattern():
    frequencies = defaultdict(Counter)

    for line in syslog:
        if re.search(r"regexforhostname",line):
            hostname= line.strip()[16:27]
            frequencies[hostname].update(f for f in word if f in line)

    return frequencies

result = findpattern()
for device, frequencies in result.iteritems():
    print '{}: {}'.format(
        device, 
        ', '.join(['{}x {}'.format(c, f) for f, c in frequencies.most_common()]))