def cut_topk(self, sentence, topk):
if not isinstance(sentence, unicode):
try:
sentence = sentence.decode('utf-8')
except UnicodeDecodeError:
sentence = sentence.decode('gbk','ignore')
graph = self.get_graph(sentence)
items = ksp_yen(graph, 0, graph.N-1, topk)
for item in items:
words = []
path = item['path']
for i in xrange(1, len(path), 1):
words.append(sentence[path[i-1]:path[i]])
yield words
def cut_topk(self, sentence, topk):
if not isinstance(sentence, unicode):
try:
sentence = sentence.decode('utf-8')
except UnicodeDecodeError:
sentence = sentence.decode('gbk', 'ignore')
graph = self.get_graph(sentence)
items = ksp_yen(graph, 0, graph.N - 1, topk)
for item in items:
words = []
path = item['path']
for i in xrange(1, len(path), 1):
words.append(sentence[path[i - 1]:path[i]])
yield words
def __cut_graph_topk(self, sentence, graph):
items = ksp_yen(graph, 0, graph.N-1, self.topk)
return self.choise_best(sentence, graph, items)
def __cut_graph_topk(self, sentence, graph):
items = ksp_yen(graph, 0, graph.N - 1, self.topk)
return self.choise_best(sentence, graph, items)
