Mypy fixes

Author: pamelafox

app/backend/prepdocslib/embeddings.py

@@ -255,15 +255,10 @@ async def create_embedding_for_image(self, image_bytes: bytes) -> list[float]:
         raise ValueError("Failed to get image embedding after multiple retries.")
 
     async def create_embedding_for_text(self, q: str):
-        if not self.endpoint:
-            raise ValueError("Azure AI Vision endpoint must be set to compute image embedding.")
         endpoint = urljoin(self.endpoint, "computervision/retrieval:vectorizeText")
         headers = {"Content-Type": "application/json"}
         params = {"api-version": "2024-02-01", "model-version": "2023-04-15"}
         data = {"text": q}
-
-        if not self.token_provider:
-            raise ValueError("Azure AI Vision token provider must be set to compute image embedding.")
         headers["Authorization"] = "Bearer " + await self.token_provider()
 
         async with aiohttp.ClientSession() as session:

app/backend/prepdocslib/listfilestrategy.py

@@ -28,10 +28,33 @@ def __init__(self, content: IO, acls: Optional[dict[str, list]] = None, url: Opt
         self.acls = acls or {}
         self.url = url
 
-    def filename(self):
-        if not self.content.name or self.content.name == "file":
-            return os.path.basename(self.content.filename)
-        return os.path.basename(self.content.name)
+    def filename(self) -> str:
+        """
+        Get the filename from the content object.
+
+        Different file-like objects store the filename in different attributes:
+        - File objects from open() have a .name attribute
+        - HTTP uploaded files (werkzeug.datastructures.FileStorage) have a .filename attribute
+
+        Returns:
+            str: The basename of the file
+        """
+        content_name = None
+
+        # Try to get filename attribute (common for HTTP uploaded files)
+        if hasattr(self.content, "filename"):
+            content_name = getattr(self.content, "filename")
+            if content_name:
+                return os.path.basename(content_name)
+
+        # Try to get name attribute (common for file objects from open())
+        if hasattr(self.content, "name"):
+            content_name = getattr(self.content, "name")
+            if content_name and content_name != "file":
+                return os.path.basename(content_name)
+
+        # If we couldn't determine a name, return a default
+        return "unknown"
 
     def file_extension(self):
         return os.path.splitext(self.filename())[1]