Bladeren bron

add tests folder

chenhaiyang 5 jaren geleden
bovenliggende
commit
d1335407cb
4 gewijzigde bestanden met toevoegingen van 16 en 7 verwijderingen
  1. 4 6
      sim/text_sim.py
  2. 0 1
      sim/words_sim.py
  3. 0 0
      tests/__init__.py
  4. 12 0
      tests/test_text_sim.py

+ 4 - 6
sim/text_sim.py

@@ -1,19 +1,17 @@
 # -*- coding: utf-8 -*-
 
-import os
+import pkgutil
+from io import StringIO
 from operator import itemgetter
 
 import jieba
 import numpy as np
-
 from .words_sim import SimCilin
 
-base_path = os.path.abspath(__file__)
-folder = os.path.dirname(base_path)
-data_path = os.path.join(folder, 'cilin_words.txt')
+data = pkgutil.get_data(__package__, 'cilin_words.txt')
+jieba.load_userdict(StringIO(data.decode()))
 
 ci_lin = SimCilin()
-jieba.load_userdict(data_path)
 
 
 def segmentation(sentence_list):

+ 0 - 1
sim/words_sim.py

@@ -12,7 +12,6 @@ data_path = os.path.join(folder, 'cilin_dict.txt')
 class SimCilin(object):
 
     def __init__(self):
-        # self.cilin_path = '/Users/highing/code/sagacloud/cilin/sim/sim/cilin_dict.txt'
         self.cilin_path = data_path
         self.sem_dict = self.load_semantic()
 

+ 0 - 0
tests/__init__.py


+ 12 - 0
tests/test_text_sim.py

@@ -0,0 +1,12 @@
+# -*- coding: utf-8 -*-
+
+"""Test for text sim."""
+
+from sim.text_sim import most_similar_items
+
+
+if __name__ == '__main__':
+
+    str1 = '我喜欢吃苹果'
+    str_l = ['我喜欢吃梨', '你喜欢吃苹果', '他喜欢吃橙子']
+    print(most_similar_items(str1, str_l, 5))