Skip to main content

Using The Extractor

To modify the extractor query, or change how it uses the local database, check out twitter-task.js.

The query object passes the fundamental instructions to our extractor.

let searchTerm = "#koii";
let query = {
    limit: 100, // total number of records to return
    searchTerm: searchTerm, // the keyword to look for
    query: `https://twitter.com/search?q=${ searchTerm }&src=typed_query`, // the query string (including said keyword)
    depth: 3, // the depth of recursive layers to follow
    recursive: true, // descend recursively?
    updateRound: () => {} // a function that returns the current round
    round: 1 // the current round
}

Testing the Extractor​

Within the tests folder, you will find multiple pre-written tests. By utilizing these, you can see how your extractor will behave in deployment.

One of the files that you will find useful is the test-one-round.js. This test will locally simulate running the extractor for one round and will output the results in the console.

You can run it with:

node test/test-one-round.js

Interacting with Task Runners​

Within config-task.yaml, you can edit how your task is presented to potential task-runners.

Environment Variables:​

Your task runners will populate this via Koii Node. We need the following:

  • Twitter Username
  • Twitter Password

Task Id:​

Do not provide a task_id for the first deployment, as it will be assigned automatically when you create your task.

Other Variables:​

Don't forget to edit the task_name, task_description, total_bounty_amount, and bounty_amount_per_round before deployment, as these would be directly visible to potential task runners in Koii Node.