raise error on encrypted files

knowledge_gpt/main.py

@@ -7,6 +7,7 @@
     is_query_valid,
     is_file_valid,
     is_open_ai_key_valid,
+    display_file_read_error,
 )
 
 from knowledge_gpt.core.caching import bootstrap_caching
@@ -43,8 +44,11 @@
 if not uploaded_file:
     st.stop()
 
+try:
+    file = read_file(uploaded_file)
+except Exception as e:
+    display_file_read_error(e)
 
-file = read_file(uploaded_file)
 chunked_file = chunk_file(file, chunk_size=300, chunk_overlap=0)
 
 if not is_file_valid(file):

knowledge_gpt/ui.py

@@ -4,6 +4,7 @@
 from knowledge_gpt.core.parsing import File
 import openai
 from streamlit.logger import get_logger
+from typing import NoReturn
 
 logger = get_logger(__name__)
 
@@ -26,15 +27,18 @@ def is_query_valid(query: str) -> bool:
 
 def is_file_valid(file: File) -> bool:
     if len(file.docs) == 0 or len(file.docs[0].page_content.strip()) == 0:
-        st.error(
-            "Cannot read document! Make sure the document has"
-            " selectable text or is not password protected."
-        )
+        st.error("Cannot read document! Make sure the document has selectable text")
         logger.error("Cannot read document")
         return False
     return True
 
 
+def display_file_read_error(e: Exception) -> NoReturn:
+    st.error("Error reading file. Make sure the file is not corrupted or encrypted")
+    logger.error(f"{e.__class__.__name__}: {e}")
+    st.stop()
+
+
 @st.cache_data(show_spinner=False)
 def is_open_ai_key_valid(openai_api_key) -> bool:
     if not openai_api_key:

pyproject.toml

@@ -19,6 +19,7 @@ pillow = "^9.4.0"
 pypdf = "^3.3.0"
 tenacity = "^8.2.0"
 tiktoken = "^0.4.0"
+pycryptodome = "^3.18.0"
 
 
 [tool.poetry.group.dev.dependencies]