Update app.py

app.py

@@ -29,6 +29,16 @@ class SanskritMorphemeTokenizer:
         """         
         if not word:             
             return [self.unk_token] 
+        if word in self.morpheme_vocab:
+            return [word]
+
+        for i in range(len(word), 0, -1):
+            prefix = word[:i]
+            if prefix in self.morpheme_vocab:
+                remaining = word[i:]
+                return [prefix] + (self._segment_word(remaining) if remaining else [])
+        return [self.unk_token]
+
  
         from functools import lru_cache