Open nj1973 opened 1 month ago
Related to this, another inconvenience around CSV headers is if we validate multiple tables in the same command and redirect the output to a CSV file then we get the header multiple times, part way through the file.
Ideally we would output the CSV header for the first run id and not for subsequent run ids.
A request from a customer:
"Currently the schema validation outputs blank entries for tables matching (with --filter-status=fail). Is it possible to suppress the table outputs in such cases?"
This would apply to all validation types. The customer is redirecting results to CSV and not storing them in BigQuery.
I've raised this on their behalf but we would need to think carefully about how this might effect other users or general usability.