It's possible you don't actually need such a key: if you uploaded the file directly, I assume you know the bucket name and object name/path? With this you can serve the object by constructing a URL (e.g. http://storage.googleapis.com/<bucket-name>/<object-path>
) and you can read it using the Cloud Storage Client Library's open()
function.
I've not used blob_keys myself, but the use case for them seems to be where an uploaded file needs to be served from a location which is unknown and therefore a GCS URL cannot be formed.