Fix for #621

corpustools/corpus/classes/lexicon.py

@@ -1060,6 +1060,11 @@ def update(self, old_word):
             else:
                 setattr(self, attribute, default_value)
 
+        if hasattr(old_word, 'wordtokens'):
+            self.wordtokens = list()
+            for wt in old_word.wordtokens:
+                self.wordtokens.append(copy.copy(wt))
+
         self.descriptors.extend([att for att in Word.word_attributes if not att.startswith('_')])
         self.descriptors = list(set(self.descriptors))
 
@@ -2088,8 +2093,9 @@ def __iadd__(self, other):
                 sw.frequency += w.frequency
                 for a in self.attributes:
                     if getattr(sw, a.name) == a.default_value and getattr(w, a.name) != a.default_value:
-                        setattr(sw, a.name, getattr(w, a.name))
-                sw.wordtokens += w.wordtokens
+                        setattr(sw, a.name, copy.copy(getattr(w, a.name)))
+                for wt in w.wordtokens:
+                    sw.wordtokens.append(copy.copy(wt))
             except KeyError:
                 self.add_word(copy.copy(w))
         if self.specifier is None and other.specifier is not None:

corpustools/corpus/classes/spontaneous.py

@@ -189,6 +189,7 @@ def keys(self):
     def __len__(self):
         return len(self.words.keys())
 
+
     def __eq__(self, other):
         if not isinstance(other,Discourse):
             return False
@@ -426,7 +427,14 @@ class WordToken():
 
 
     """
-    def __init__(self,**kwargs):
+    wordtoken_attributes = {'wordtype': None, 'discourse': None, 'speaker': None, 'wavpath': None, 'begin': None,
+                            'end': None, '_spelling': None, '_transcription': None, '_freq_names': None}
+    def __init__(self,update=False,**kwargs):
+
+        if update:
+            self.update(update)
+            return
+
         self.wordtype = kwargs.pop('word',None)
         self.discourse = None
         self.speaker = None
@@ -493,6 +501,20 @@ def __init__(self,**kwargs):
                 else:
                     self._spelling = None
 
+    def __copy__(self):
+        return WordToken(update=self)
+
+    def update(self, old_token):
+
+        for attribute, value in old_token.__dict__.items():
+            if not hasattr(self, attribute):
+                setattr(self, attribute, value)
+
+        for attribute, default_value in WordToken.wordtoken_attributes.items():
+            if hasattr(old_token, attribute):
+                setattr(self, attribute, getattr(old_token, attribute))
+            else:
+                setattr(self, attribute, default_value)
 
     def __getstate__(self):
         state = self.__dict__.copy()