Update data_processor.py

data_processor.py

@@ -30,12 +30,16 @@ try:
 except:
     pass
 """
-try:
-    from nltk.corpus import stopwords
-    _ = stopwords.words("english")
-except LookupError:
-    nltk.download("stopwords")
-    from nltk.corpus import stopwords
+resources = ["stopwords", "punkt", "punkt_tab"]
+for res in resources:
+    try:
+        if res == "stopwords":
+            from nltk.corpus import stopwords
+            _ = stopwords.words("english")
+        elif res in ["punkt", "punkt_tab"]:
+            nltk.data.find(f"tokenizers/{res}")
+    except LookupError:
+        nltk.download(res)
 
 class FinancialDataProcessor:
     """Processes financial documents and generates Q&A pairs"""