Merge pull request #919 from VesnaT/language_unpickle

[FIX] Corpus: Unpickle corpus without language

orangecontrib/text/corpus.py

@@ -4,7 +4,7 @@
 from copy import copy, deepcopy
 from numbers import Integral
 from itertools import chain
-from typing import Union, Optional, List, Tuple
+from typing import Union, Optional, List, Tuple, Dict
 from warnings import warn
 
 import nltk
@@ -368,6 +368,11 @@ def titles(self):
     def language(self):
         return self.attributes["language"]
 
+    def __setstate__(self, state: Dict):
+        super().__setstate__(state)
+        if "language" not in self.attributes:
+            self.attributes["language"] = None
+
     def documents_from_features(self, feats):
         """
         Args:

orangecontrib/text/tests/test_corpus.py

@@ -691,6 +691,12 @@ def test_remove_attributes_copy(self):
         """
         self.assertLess(datetime.today(), datetime(2024, 1, 1))
 
+    def test_language_unpickle(self):
+        path = os.path.dirname(__file__)
+        file = os.path.abspath(os.path.join(path, "data", "book-excerpts.pkl"))
+        corpus = Corpus.from_file(file)
+        self.assertIsNone(corpus.attributes["language"])
+
 
 @skipIf(summarize is None, "summarize is not available for orange3<=3.28")
 class TestCorpusSummaries(unittest.TestCase):

orangecontrib/text/widgets/tests/test_owcorpus.py

@@ -1,3 +1,4 @@
+import os
 import unittest
 
 import numpy as np
@@ -325,6 +326,15 @@ def test_guess_language(self):
         self.assertEqual("English", self.widget.language)
         self.assertEqual("en", self.get_output(self.widget.Outputs.corpus).language)
 
+    def test_language_unpickle(self):
+        path = os.path.dirname(__file__)
+        file = os.path.abspath(os.path.join(path, "..", "..", "tests",
+                                            "data", "book-excerpts.pkl"))
+        corpus = Corpus.from_file(file)
+        self.send_signal(self.widget.Inputs.data, corpus)
+        self.wait_until_finished()
+        self.assertEqual(self.widget.language, "English")
+
 
 if __name__ == "__main__":
-    unittest.main()
+    unittest.main()