set = hash set constructed by counting each occurance of W
sum_w = len(W), used to count the total characters needed for an anagram. 0 when substring is an anagram
sum_extra = 0, used to count extra characters not part of the anagram

anagram_idxs = empty linked list

substring = S [1:len (W)] <- first part of the string
for each character c in W:
	if c is in set d
	   	sum -= 1
	else:
		sum_extra += 1


if sum == 0 and sum_extra == 0:
   then the initial substring is an anagram
   anagram_idx.append(1) or 0, depending on if you want index or character location


^ up to here we've used up to O(len(W)) time.

for i from 2 to len(S) - len(W): <- I might be one off here
	missing_character = S[i-1]

	if missing_character is in set:
	   sum_w += 1
	else:
	   sum_extra -= 1


	new_character = S[i+len(W)] <- might be one off

	if new_character is in set:
	   sum -= 1
	else:
		sum_extra += 1

	if sum == 0 and sum_extra == 0:
	   anagram_idxs.append(i)