Improve the camel-quarkus-google-bigquery to test with a large table

### Describe the feature here

Currently we lack a test for a large table, it could use ARROW data format as a stream result. Avoiding storing huge files for the stubbed data in the repo, we perhaps generate those at build time.

Start by locating the camel-quarkus-google-bigquery component and its existing tests. Review how stubbed table data is provided, then define a large-table case using ARROW stream results and generated build-time data; done means the test exercises a large table without committing huge files.

Written by the indexing model from the issue text.