A program for periodically updating Wilt user's play histories.
This is intended to run on AWS Lambda using CloudWatch to trigger updates.
swift testTo build a packaged zip that contains the built program, run:
./deploy.shThis will create a zip called WiltCollector.zip in deploy/.
This can then be uploaded to S3 and connected to Lambda. index.js is the
Lambda handler and currently logs all output from the Swift program and the
handler returns an empty string on completion.
You'll need to copy your credentials file at $GOOGLE_APPLICATION_CREDENTIALS
to deploy/credentials.json.
You'll need to set environment variables BIGQUERY_PROJECT_ID,
SPOTIFY_CLIENT_ID and SPOTIFY_CLIENT_SECRET in the Lambda console.
NOTE: BIGQUERY_PROJECT_ID is now also used for the Firebase project ID,
so you'll need to ensure these exist under the same project within Google
Cloud. These should be separated into separate environment variables at some
point soon.
You'll need to create a BigQuery table called wilt_play_history.play_history.
You'll need to create a table called users.
play_history (BigQuery): user_id, date, artists, name, primary_artist, track_id
users (FireStore): access_token, expires_at, refresh_token
- Some sort of exponential backoff if requests start failing (in case Spotify bans me)