Trigger a Dataprep job when a file is uploaded to Cloud Storage using a Cloud Function. This works by calling the Dataprep API.

For triggering Dataprep flows using Dataflow templates directly rather than the Dataprep API, see gcp-dataflow-gcf-trigger.

Run the following command to deploy this Cloud Function:

$ gcloud functions deploy <function-name> \
    --entry-point trigger \
    --trigger-bucket gs://<my-bucket> \
    --set-env-vars RECIPE_ID=<dataprep-recipe>,DATAPREP_TOKEN=<dataprep-token> \
    --runtime python37

Ensure that the trigger bucket matches the bucket used for the Dataprep recipe.

As shown above, this Cloud Function requires two environment variables:

• RECIPE_ID: the ID of the Dataprep recipe to start jobs for. The provided recipe must have a parameterized dataset with a variable called file which is the file to process from the configured bucket.
• DATAPREP_TOKEN: Dataprep access token used to authenticate with the Dataprep API