Merge pull request #447 from s-alexey/pickle_udpipe

Pickling UDPipe #387
biolab · Aug 27, 2019 · 70be2b8 · 70be2b8
2 parents 49c24a6 + 0b678f3
commit 70be2b8
Show file tree

Hide file tree

Showing 2 changed files with 17 additions and 0 deletions.
diff --git a/orangecontrib/text/preprocess/normalize.py b/orangecontrib/text/preprocess/normalize.py
@@ -188,3 +188,10 @@ def language(self):
     def language(self, value):
         self._language = value
         self.model = None
+
+    def __getstate__(self):
+        return {'language': self.language}
+
+    def __setstate__(self, state):
+        self.__init__(state['language'])
+
diff --git a/orangecontrib/text/tests/test_preprocess.py b/orangecontrib/text/tests/test_preprocess.py
@@ -1,3 +1,4 @@
+import pickle
 import tempfile
 import unittest
 import os.path
@@ -187,6 +188,15 @@ def test_udpipe_doc(self):
         self.assertListEqual(normalizer.normalize_doc('Gori na gori hiša gori'),
                              ['gora', 'na', 'gora', 'hiša', 'goreti'])
 
+    def test_udpipe_pickle(self):
+        normalizer = preprocess.UDPipeLemmatizer()
+        normalizer.language = 'English'
+
+        loaded = pickle.loads(pickle.dumps(normalizer))
+        self.assertEqual(normalizer.language, loaded.language)
+        self.assertEqual(loaded.normalize_doc('peter piper pickled'),
+                         ['peter', 'piper', 'pickle'])
+
     def test_porter_with_bad_input(self):
         stemmer = preprocess.PorterStemmer()
         self.assertRaises(TypeError, stemmer, 10)