Is there any way to get the number of rows actually loaded, but not the total count for a table.

[zhongdai]

Here is an example, the code snippet is

    # make a call of loadjob and use .result() to wait until it competes

    table = client.get_table(table_id)  # Make an API request.
    print(
        "Loaded {} rows and {} columns to {}".format(
            table.num_rows, len(table.schema), table_id
        )
    )

After the load job completed, it calls get_table to get the num_rows, and prints it. It works if we do a fresh load to an empty table, but when we load a file (or files) to an existing table by using APPEND mode, that number of rows is not the actually loaded count.

My workaround is, get_table() twice and calculate the difference between two counts. I am wondering is there any better approach to get the actually loaded rows (it is like the affected rows in other DBMS).

[mehul1811]

I am facing the same issue when appending the results of a query to an existing table using the QUERY job. Below code also gives the total count of rows in the destination table after appending:

query_job_config = bigquery.QueryJobConfig(destination_table)
query_job = bq_client.query(sql, job_config=query_job_config)
results = query_job.result()
print(results.total_rows)

[chalmerlowe]

Based on staffing and project priorities, this item will not be worked.