Hi folks, I am using the Python Avro bindings (both 2 and 3), and I have a schema which has a top-level union comprised of different record types. I am able to construct records, wrap them in the union, and serialize them just fine. I then throw the serialized union into a pile for a consumer to grab.
When my consumer grabs a serialized union from the pile, I deserialize it and I want to begin processing the data. The first thing it wants to do is determine the record type so it can figure out how to process the data. I didn't see a way to actually determine the record type of the selected union branch in the python bindings. Am I missing something obvious? http://search-hadoop.com/m/F2svI1fnW971bMJRh seems to have some workarounds, but I was wondering if there was a preferred way of handling this scenario, or if there is a better approach to processing the deserialized record that I am missing. Thanks, -- Tim Upthegrove