Fix docs parser (#1714)

### What problem does this PR solve?

#1711 

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)

rag/nlp/__init__.py

@@ -550,7 +550,7 @@ def concat_img(img1, img2):
 
 def naive_merge_docx(sections, chunk_token_num=128, delimiter="\n。；！？"):
     if not sections:
-        return []
+        return [], []
 
     cks = [""]
     images = [None]