Skip to content
GeoJibe

Import from Google BigQuery

Use Google BigQuery when the data already lives in a BigQuery table or view.

You need to be signed in. Saving the connection once is the usual path.

This is vector table import. GeoJibe copies one table into the workflow; it is not a database migration, SQL editor, or BigQuery administration tool.

You do not write SQL for an ordinary table import. Choose Project → Dataset → Table.

Before you start

You'll need:

  • A Google Cloud project ID
  • A service-account JSON key with permission to read the dataset
  • A BigQuery dataset and table or view

Typical IAM roles for import:

  • roles/bigquery.jobUser on the project (so GeoJibe can run a read job)
  • roles/bigquery.dataViewer on the dataset (or the table)

The saved Source is the project and service account. You choose the dataset and table each time you import or save a Job.

Paste or upload the JSON key in the GeoJibe UI. You do not set Google credentials as environment variables on the GeoJibe server.

1. Save the Google BigQuery Source

Open Sources → Add Source → Google BigQuery.

Fill in Project and choose the service-account JSON key file. Click Test, then Save. Name it something like Analytics BigQuery.

Editing a saved Source without choosing a new key file keeps the stored credential.

2. Add Google BigQuery on the Canvas

Add Google BigQuery, choose the saved Source or enter the project and key, then Connect.

Choose a dataset, then a table or view. Tables with a GEOGRAPHY column are marked. Click Import. GeoJibe streams the table into the job workspace and inspects it.

After Import, Transform sees the real field names. Convert offers the same formats as an uploaded file.

You can also fill the connection in manually for a one-off run. You do not have to save a Job.

GEOGRAPHY and CRS

BigQuery GEOGRAPHY is WGS 84 longitude/latitude (EPSG:4326). GeoJibe uses the first GEOGRAPHY column as feature geometry. Extra GEOGRAPHY columns become WKT text.

Nested RECORD / STRUCT fields and REPEATED arrays are not imported in this version. GeoJibe reports those tables instead of silently flattening them.

NUMERIC and BYTES values are imported as text so precision is not lost.

3. Process and deliver

Add Transform, Convert, or Python if you need them. Then choose a Destination — Download, PostGIS, another BigQuery table, and the rest work as usual.

Example: Google BigQuery → Transform → PostGIS.

Click Run.

See also