A few years back, when working on SQL Server projects, I often utilized the ALTER TABLE SWITCH partition between staging and target tables. This left me pondering—could a similar functionality be achieved in BigQuery? 🤔

BigQuery facilitates copying a partition to another table using the bq cp command:

bq cp -f 'project:dataset.source_table$your_partition' 'project:dataset.target_table$your_partition'

🧪 Example Scenario 🧪

Imagine a staging table, where we extract the delta from a source table (based on the last entry seen in the target table), and MERGE it into the target table.

Instead, we can craft a short script to copy delta partitions (new + changed, if any) into the target table, bypassing the need for MERGE altogether! This generates COPY jobs as opposed to QUERY jobs.

#!/bin/bash

# Check if the correct number of arguments is provided
if [[ "$#" -ne 2 ]]; then
    echo "Usage: $0 <start_date> <end_date>"
    exit 1
fi

# Define the project and dataset names
PROJECT="***********"
DATASET_SOURCE="learning.data_source_staging"
DATASET_DEST="learning.data_source"

# Assign start date and end date from input arguments
START_DATE=$1
END_DATE=$2

# Loop through the dates from START_DATE to END_DATE
for date in $(seq -w $START_DATE $END_DATE); do
  # Echo a message indicating the partition being swapped
  echo "Swapping partition $date"

  # Run the bq cp statement
  bq cp -f "${PROJECT}:${DATASET_SOURCE}\$$date" "${PROJECT}:${DATASET_DEST}\$$date"
done

Once this executes, we will have copied the partitions from the staging table to the target table using the bq cp command.

We can confirm that all the partitions have been loaded properly and see the list of COPY jobs in the job history tab.

Also, don’t forget—leveraging the INFORMATION_SCHEMA PARTITIONS view can assist in constructing even more advanced functionalities.

Found it useful? Subscribe to my Analytics newsletter at notjustsql.com.