Incomplete BigQuery schema on Nested JSON History jobs?

Question

The schema on a job of BigQuery for a Nested Json file, does not keep the original schema submitted on the import job. If you look the table schema under dataset, they match correctly.

But if you try to Repeat load job under the Job History of BigQuery Web Interface it shows only the first level of the schema, preventing to re run the same job. Since the job schema is different than the table schema.

Sample schema on the Job not showing the actual fields:

Header: RECORD Line: RECORD MetaData: RECORD

Is this behavior as design?

Answer 1

This is bug. I've filed it internally, we should hopefully have a fix soon. It isn't actually a question of keeping the schema, it is a question of translating it into the 'inline schema' format. If you paste the json schema into the field, it should work.