tap-outbrain

Author: Connor McArthur (connor@fishtownanalytics.com)

Development

This repository contains two configuration files that need to be copied and filled out before running:

• config.json.example: copy to config.json in the repo root. Contains:

  • account_id, aka the Marketer ID (unique to each account) in Outbrain. Looks like 00f4b02153ee75f3c9dc4fc128ab041962.
  • username, the Outbrain username used to generate an Amplify API token.
  • password, the Outbrain password to go along with username.
  • access_token, an optional argument. If provided, this will be used as the access token, and a new one won't be generated.

• persist.json.example: copy to persist.json in the repo root. Contains the configuration for the Stitch persister.

Then, the streamer can be run (with persistence) with:

docker build -f Dockerfile .
docker run <image-id>

Or, for interactive development:

docker build -f Dockerfile .

# this automatically propagates changes into the container
docker -v "$(pwd)":/usr/src/tap-outbrain run <image-id>

Gotchas

• Outbrain only allows two calls to the /login API per hour. This integration calls that API on every run to generate a new access token. This means that this integration cannot be run more frequently than twice per hour. The access token could be stored in the state file with a timestamp, but at present secure state file storage is not implemented.
• Campaign pagination is not implemented -- this integration pulls incomplete data if more than 100 campaigns exist.

---

Copyright © 2017 Fishtown Analytics