Best way to count the number of occurrences of a substring?

Question

C = 'Kabansososkabansosos'

I need the number of times the string 'sos' occurs in C. I've already used the C.count('sos') method, but it did not help me.

Answer 1

def count_occurrences(expression,word): 
    L = len(word)
    counts = 0
    for i in range(L,len(expression)+1):
        if expression[i-L:i]==word:
            counts += 1
    return counts

count_occurrences('Kabansososkabansosos','sos')
>>> 4

Answer 2

To be explicit, your problem is that you want to count non-overlapping occurrences of the sub-string, which is not what str.count() does.

First, let's use the powerful re (regular expression) module instead of str.count():

C = 'Kabansososkabansosos'
matches = re.findall(r'sos', C)
n = len(matches)
print(n)  # 2

At first sight this seems silly, as it 1) also does not work with over-lapping occurrences and 2) produce a (potential) large list of matches, all of which will just be the str 'sos'.

However, using the clever trick of the accepted answer here, we can make re.findall() work with overlapping occurrences:

C = 'Kabansososkabansosos'
matches = re.findall(r'(?=(sos))', C)
n = len(matches)
print(n)  # 4

If C is very large, we would like to avoid the potentially large temporary list matches. We can do this by instead using the re.finditer() function, which returns a generator rather than the list:

C = 'Kabansososkabansosos'
matches = re.finditer(r'(?=(sos))', C)
n = sum(1 for match in matches)
print(n)  # 4

where the sum() over 1s simply produce the length of the iterator, i.e. the number of matches.