Merge pull request #43 from mindvalley/feat/parse-google-spreadsheet

Update the logic to parse Google Spreadsheet in Gdrive connector
mindvalley · Jun 20, 2024 · 04b86b3 · 04b86b3
2 parents 5757a58 + 6e69ffb
commit 04b86b3
Showing 1 changed file with 4 additions and 1 deletion.
diff --git a/backend/danswer/connectors/google_drive/connector.py b/backend/danswer/connectors/google_drive/connector.py
@@ -334,7 +334,10 @@ def extract_text(file: dict[str, str], service: discovery.Resource) -> str:
         response = service.files().get_media(fileId=file["id"]).execute()
         return pptx_to_text(file=io.BytesIO(response))
     elif mime_type == GDriveMimeType.PPT.value:
-        response = service.files().get_media(fileId=file["id"]).execute()
+        response = service.files().export(
+            fileId=file["id"],
+            mimeType="application/vnd.openxmlformats-officedocument.presentationml.presentation"
+        ).execute()
         return pptx_to_text(file=io.BytesIO(response))
 
     return UNSUPPORTED_FILE_TYPE_CONTENT