def test_text():
cg = CorpusGraph()
# 从json文件读取语料库模型
# cg.load_from_json()
# 连接mongodb建立语料库模型
cg.build_corpus()
# 保存为json文件
cg.save_as_json()
tg = TextGraph()
# 从mongodb读取句子，以便分词
# sentences = tg.get_sentences(isRandom=False)
sentences = ["准许原告肖振明撤回起诉"]
# 对句子数组建立图模型
tg.build(sentences)
# 填入边的权重
tg.fill_edge(cg)
# 输出语句图需要的json文件, path如果为None则返回json，而不保存在硬盘
tg.make_json(cg, path='./data/text.json')
def test_text():
cg = CorpusGraph()
cg.build_corpus()
cg.get_sorted_neighbour('一')
# print("###############")
# for cge in cg.corpus.edges:
# print(cge)
# break
# print('###', cg.corpus['朝'])
tg = TextGraph()
sentences = tg.get_sentences(isRandom=False)
tg.build(sentences)
tg.fill_edge(cg)
tg.make_json(cg)
def tokenize():
if request.method == 'GET':
tg = TextGraph()
sentence = "没有输入"
# 从参数获取待分词句子
if request.args.get('sentence', '') != "":
sentence = request.args.get('sentence', '')
tg.build([sentence])
tg.fill_edge(cg)
# 暂时只对单句分词
time_count(print_to_console=False)
result = tg.cut()[0]
time_count("分词完毕")
check_jieba = jieba_checker.check(sentence, result)
time_count("jieba分词完毕")
check_thulac = thulac_checker.check(sentence, result)
time_count("thulac分词完毕")
# jieba的分词结果
jieba_result = check_jieba["jieba_result"]
jieba_overlap = check_jieba["overlap"]
thulac_result = check_thulac["thulac_result"]
thulac_overlap = check_thulac["overlap"]
# res = json.dumps(
# {"graph": tg.make_json(cg, path=None), "result": result,
# "jieba": jieba_result, "jieba_overlap": jieba_overlap,
# "thulac": thulac_result, "thulac_overlap": thulac_overlap},
# ensure_ascii=False)
res = json.dumps(
{
"graph": tg.make_json(cg, path=None),
"result": result,
"jieba": {
"words": jieba_result,
"overlap": "%.2f" % jieba_overlap
},
"thulac": {
"words": thulac_result,
"overlap": "%.2f" % thulac_overlap
}
},
ensure_ascii=False)
# print("json dumping")
# res = json.dumps(
# {"graph": tg.make_json(cg, path=None), "result": result,
# "jieba": jieba_result, "jieba_overlap": jieba_overlap,
# },
# ensure_ascii=False)
print("server returned")
return res
def tokenize(sentence):
tg = TextGraph()
tg.build([sentence])
tg.fill_edge(cg)
# 暂时只对单句分词
result = tg.cut()[0]
check = checker.check(sentence, result)
jieba_result = check["jieba_result"]
overlap = check["overlap"]
res = json.dumps(
{"result": result, "jieba": jieba_result, "overlap": overlap},
ensure_ascii=False)
return res
def tokenize(sentence):
tg = TextGraph()
tg.build([sentence])
tg.fill_edge(cg)
# 暂时只对单句分词
result = tg.cut()[0]
jieba_check = jieba_checker.check(sentence, result)
thulac_check = thulac_checker.check(sentence, result)
jieba_result = jieba_check["jieba_result"]
jieba_overlap = jieba_check["overlap"]
thulac_result = thulac_check["thulac_result"]
thulac_overlap = thulac_check["overlap"]
res = {"sentence": sentence, "result": result, "jieba": jieba_result, "jieba_overlap": jieba_overlap,"thulac":thulac_result,"thulac_overlap":thulac_overlap}
return res