Merge pull request #594 from NatLibFi/upgrade-simplemma-0.7

Upgrade Simplemma to version 0.7

annif/analyzer/simplemma.py

@@ -10,17 +10,8 @@ class SimplemmaAnalyzer(analyzer.Analyzer):
 
     def __init__(self, param, **kwargs):
         self.lang = param
-        self.langdata = None
         super().__init__(**kwargs)
 
-    def __getstate__(self):
-        """Return the state of the object for pickling purposes. The langdata
-        field is set to None as it's more efficient to use load_data."""
-
-        return {'lang': self.lang, 'langdata': None}
-
     @functools.lru_cache(maxsize=500000)
     def _normalize_word(self, word):
-        if self.langdata is None:
-            self.langdata = simplemma.load_data(self.lang)
-        return simplemma.lemmatize(word, self.langdata)
+        return simplemma.lemmatize(word, self.lang)

setup.py

@@ -42,7 +42,7 @@ def read(fname):
         'stwfsapy==0.3.*',
         'python-dateutil',
         'tomli==2.0.*',
-        'simplemma==0.6.*'
+        'simplemma==0.7.*'
     ],
     tests_require=['py', 'pytest', 'requests'],
     extras_require={

tests/test_analyzer_simplemma.py

@@ -6,12 +6,6 @@
 simplemma = pytest.importorskip("annif.analyzer.simplemma")
 
 
-def test_simplemma_getstate():
-    analyzer = annif.analyzer.get_analyzer("simplemma(fi)")
-    state = analyzer.__getstate__()
-    assert state == {'lang': 'fi', 'langdata': None}
-
-
 def test_simplemma_finnish_analyzer_normalize_word():
     analyzer = annif.analyzer.get_analyzer("simplemma(fi)")
     assert analyzer._normalize_word("xyzzy") == "xyzzy"