bug12

app.py

@@ -16,8 +16,8 @@ def pdf_to_text_with_layout(pdf_file):
     return "\n".join(text)
 
 def clean_text(text):
-    # Remove non-XML-compatible characters
-    return re.sub(r'[^\x09\x0A\x0D\x20-\xD7FF\xE000-\xFFFD\x10000-\x10FFFF]', '', text)
+    # Remove non-ASCII and control characters
+    return ''.join(c for c in text if c.isprintable() and ord(c) < 65536)
 
 def text_to_word_with_formatting(text, word_path):
     doc = Document()