I have a Dataflow pipeline that reads from a BigQuery table. However, when reading the data, there is no other option than to read all records with the read(SerializableFunction) or the readTableRows() methods. I was wondering, when using these methods, is there an option to provide a column-like projection (similar to the HBaseIO Scan addColumn() filter).
I know that there is the option of performing a query when loading data from BigQuery (using the readTableRows().fromQuery()), but I'm wondering if there is a similar option to do projection like possible in HBaseIO.