My (working) code roughly does the following: read_session = bigquery_storage.ReadSession(...) request = bigquery_storage.CreateReadSessionRequest(...) session = client.create_read_session(request=request) iterator = [client.read_rows(stream.name).rows(session) for stream in session.streams][0] # and here I can nicely iterate over the rows... However, I hate about the code that it is blocking. I discovered in the docs that a BigQueryReadAsyncClient has a async create_read_session as a drop-in replacement of the same-named call in the snippet above (link to docs). Unfortunately, according to the docs, this still returns a regular ReadSession without any async methods. Am I overlooking something? Is there a nice, async way to lazily iterate over the rows of a bigquery table / query result with some async buffer? Seems like a reasonably standard use case thus I wonder what I am missing. Continue reading...
