improve handling of bad content types

open_notebook/exceptions.py

@@ -10,6 +10,12 @@ class DatabaseOperationError(OpenNotebookError):
     pass
 
 
+class UnsupportedTypeException(OpenNotebookError):
+    """Raised when an unsupported type is provided."""
+
+    pass
+
+
 class NoSchemaFound(OpenNotebookError):
     """Raised when a database schema is not found."""
 

open_notebook/graphs/content_process.py

@@ -16,6 +16,7 @@ from youtube_transcript_api import YouTubeTranscriptApi  # type: ignore
 from youtube_transcript_api.formatters import TextFormatter  # type: ignore
 
 from open_notebook.config import CONFIG
+from open_notebook.exceptions import UnsupportedTypeException
 
 
 class SourceState(TypedDict):
@@ -513,7 +514,9 @@ def file_type_edge(data: SourceState):
     elif data.get("identified_type").startswith("audio"):
         return "extract_audio"
     else:
-        return "end"
+        raise UnsupportedTypeException(
+            f"Unsupported file type: {data.get('identified_type')}"
+        )
 
 
 workflow = StateGraph(SourceState)

stream_app/source.py

@@ -9,6 +9,7 @@ from loguru import logger
 
 from open_notebook.config import UPLOADS_FOLDER
 from open_notebook.domain import Asset, Source
+from open_notebook.exceptions import UnsupportedTypeException
 from open_notebook.graphs.content_process import graph
 from open_notebook.graphs.multipattern import graph as transform_graph
 from open_notebook.utils import surreal_clean
@@ -138,17 +139,31 @@ def add_source(session_id):
         logger.debug("Adding source")
         with st.status("Processing...", expanded=True):
             st.write("Processing document...")
-            result = graph.invoke(req)
+            try:
-            st.write("Saving..")
+                result = graph.invoke(req)
-            source = Source(
+                st.write("Saving..")
-                asset=Asset(url=req.get("url"), file_path=req.get("file_path")),
+                source = Source(
-                full_text=surreal_clean(result["content"]),
+                    asset=Asset(url=req.get("url"), file_path=req.get("file_path")),
-                title=result.get("title"),
+                    full_text=surreal_clean(result["content"]),
-            )
+                    title=result.get("title"),
-            source.save()
+                )
-            source.add_to_notebook(st.session_state[session_id]["notebook"].id)
+                source.save()
-            st.write("Summarizing...")
+                source.add_to_notebook(st.session_state[session_id]["notebook"].id)
-            source.generate_toc_and_title()
+                st.write("Summarizing...")
+                source.generate_toc_and_title()
+            except UnsupportedTypeException:
+                st.warning(
+                    "This type of content is not supported yet. If you think it should be, let us know on the project Issues's page"
+                )
+                st.link_button(
+                    "Go to Github Issues",
+                    url="https://www.github.com/lfnovo/open_notebook/issues",
+                )
+                st.stop()
+
+            except Exception as e:
+                st.error(e)
+                return
 
         st.rerun()