Fix docx images (#2756)

### What problem does this PR solve?

#2755 

### Type of change

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

rag/app/naive.py

@@ -16,14 +16,15 @@ from docx import Document
 from timeit import default_timer as timer
 import re
 from deepdoc.parser.pdf_parser import PlainParser
-from rag.nlp import rag_tokenizer, naive_merge, tokenize_table, tokenize_chunks, find_codec, concat_img, naive_merge_docx, tokenize_chunks_docx
+from rag.nlp import rag_tokenizer, naive_merge, tokenize_table, tokenize_chunks, find_codec, concat_img, \
+    naive_merge_docx, tokenize_chunks_docx
 from deepdoc.parser import PdfParser, ExcelParser, DocxParser, HtmlParser, JsonParser, MarkdownParser, TxtParser
 from rag.settings import cron_logger
 from rag.utils import num_tokens_from_string
 from PIL import Image
 from functools import reduce
 from markdown import markdown
-from docx.image.exceptions import UnrecognizedImageError
+from docx.image.exceptions import UnrecognizedImageError, UnexpectedEndOfFileError, InvalidImageStreamError
 
 
 class Docx(DocxParser):
@@ -42,6 +43,12 @@ class Docx(DocxParser):
         except UnrecognizedImageError:
             print("Unrecognized image format. Skipping image.")
             return None
+        except UnexpectedEndOfFileError:
+            print("EOF was unexpectedly encountered while reading an image stream. Skipping image.")
+            return None
+        except InvalidImageStreamError:
+            print("The recognized image stream appears to be corrupted. Skipping image.")
+            return None
         try:
             image = Image.open(BytesIO(image_blob)).convert('RGB')
             return image
@@ -276,7 +283,9 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
 if __name__ == "__main__":
     import sys
 
+
     def dummy(prog=None, msg=""):
         pass
 
+
     chunk(sys.argv[1], from_page=0, to_page=10, callback=dummy)