Here's my use case: I have an item to upload with multiple files and metadata that varies per file. So I'm calling upload(id, files=files, metadata=metadata) once with the item metadata, then repeatedly calling modify_metadata(id, target="files/<filename>", metadata=<file_metadata>) for each of the files. I'd like to not start a derive until the entire process is finished. I can set queue_derive=False on the upload call, but modify_metadata doesn't take a queue_derive or headers argument, so there's no obvious way to start the derive task short of re-uploading one of the files (e.g., the smallest one).
I would also be happy if someone can suggest some no-op operation that would allow me to queue a derive as a side effect.
Here's my use case: I have an item to upload with multiple files and metadata that varies per file. So I'm calling
upload(id, files=files, metadata=metadata)once with the item metadata, then repeatedly callingmodify_metadata(id, target="files/<filename>", metadata=<file_metadata>)for each of the files. I'd like to not start a derive until the entire process is finished. I can setqueue_derive=Falseon theuploadcall, butmodify_metadatadoesn't take aqueue_deriveorheadersargument, so there's no obvious way to start the derive task short of re-uploading one of the files (e.g., the smallest one).I would also be happy if someone can suggest some no-op operation that would allow me to queue a derive as a side effect.