Details
-
Improvement
-
Status: Open
-
P3
-
Resolution: Unresolved
-
None
-
None
Description
`BigQueryIO.write` should be able to validate the schema of the input PCollection against the schema for this table (either retrieved from DataCatalog, or from BQ metadata). Then we can fail at pipeline construction time rather than waiting until execution.
At the very least this should work when writing Beam Rows.