Looker Generator

Looker Generator is a tool to generate LookML from BigQuery database schema.

Installation

• Python 3.9 or later
• make setup

Authenticating with gcloud

make auth

You don't need a service key, but your personal account might need a specific permission to access the project.

Update your env file

Create a .env file and fill in the relevant details. Use .env.example as a template.

Creating a view

1. Create a view in BigQuery in `ct-looker-*
2. Create a file in /configs with the same name as the view and fill out the relevant details
3. make run

What this does

It does a few things with a view:

1. Adds a description to the view if one exists in the table
2. Identify dimensions, dimension_groups and measures
3. Creates nested views for each array field
4. Tries to identify Primary Keys if the column is called ID, PK or primary)key
5. Light-touch labelling of potentially diagnostic or Snowplow-identifiable fields
6. Remove time from date-based dimension groups

Testing

Tests are written using pytest. To run the tests, run the following command:

make test

A coverage report will be generated in the coverage directory.