def areSentencesSimilarTwo(self, words1: List[str], words2: List[str],
pairs: List[List[str]]) -> bool:
global_incr = 0
def getWordIdx(word: str,
dictionary: Dict[str, int],
create: bool = False) -> int:
nonlocal global_incr
if word in dictionary:
return dictionary[word]
elif create:
dictionary[word] = global_incr
global_incr += 1
return dictionary[word]
return -1
if len(words1) != len(words2):
return False
N = len(pairs) * 2
uf = UF(N)
# use a word-to-index map to map all words to UnionFind's int data structure (starting from 0!)
hashmap = dict()
# build union-find forest from all similar pairs, similar words will compose a cycle inside
for edge in pairs:
u = getWordIdx(edge[0], hashmap, True)
v = getWordIdx(edge[1], hashmap, True)
uf.union(u, v)
# compare each word in words1 and words2 to see whether they are similar
for i, w1 in enumerate(words1):
w2 = words2[i]
if w1 == w2: continue
u, v = getWordIdx(w1, hashmap), getWordIdx(w2, hashmap)
if u < 0 or v < 0: return False
if uf.find(u) != uf.find(v): return False
return True
def findCircleNum(self, M: List[List[int]]) -> int:
N = len(M)
uf = UF(N)
for i in range(N):
for j in range(i + 1, N):
if M[i][j]: # merge as many direct friends as possible into clusters
uf.union(i, j)
# how many root nodes after unions
clusters = set()
for i in range(N):
clusters.add(uf.find(i))
size = len(clusters)
print(size)
return size