A small sample that listens to messages posted in a space in IBM Watson Workspace and returns back recent tweets about a given topic.
JavaScript
Switch branches/tags
Nothing to show
Latest commit ef74a42 Nov 2, 2016 @lbenitez3000 lbenitez3000 committed on GitHub Merge pull request #1 from lbenitez3000/initial
Initial commit
Permalink
Failed to load latest commit information.
.bluemix Initial commit Nov 2, 2016
.babelrc Initial commit Nov 2, 2016
.cfignore Initial commit Nov 2, 2016
.gitignore Initial commit Nov 2, 2016
.travis.yml Initial commit Nov 2, 2016
LICENSE Initial commit Nov 1, 2016
README.md Initial commit Nov 2, 2016
app.js Initial commit Nov 2, 2016
package.json Initial commit Nov 2, 2016

README.md

Twitter App Sample for IBM Watson Workspace

Build Status

App built with node.js to listen to messages posted in a space in IBM Watson Workspace and returns back recent tweets about a given topic.

This app will listen for @twitter <keyword/hashtag/@user/phrase> and give you a sense of what people are saying about that in Twitter.

The Watson Work platform provides spaces for people to exchange messages in conversations. This app shows how to listen to a conversation and receive messages on a Webhook endpoint, then send response messages back to the conversation. It also demonstrates how to authenticate an application and obtain the OAuth token needed to make Watson Work API calls.

Deploy the app

Assuming you just want to take this code and get it running before hacking it, the first step is to get it deployed to a live on a server so that IBM Watson Workspace validates it's up and working before pushing messages to the app.

To facilitate things, you can click the button below and it'll get it going to Bluemix very easily, however this is NOT required. Feel free to deploy on any server you want.

Deploy to Bluemix

Note: you can run this code locally, but then you would need to change the webhook code to get the URL of the callback. See Appendix for instructions.

Once the code is live, make note of the URL (e.g. http://mysuperawesomeurl.mybluemix.net) as you'll need this later.

How to register the Twitter App with IBM Watson Workspace

Now let's register the app to get some API keys and get things going. First, we need several API keys from Twitter. You can get them by going go to: https://apps.twitter.com . In a couple of steps you can super easily register a "bot". That will then give you a few keys that you will also need to put in your environment variables as such:

In Terminal (or any Terminal emulator such as Git Bash or Cygwin):

export TWITTER_CONSUMER_KEY=...
export TWITTER_CONSUMER_SECRET=...
export TWITTER_ACCESS_TOKEN_KEY=....
export TWITTER_ACCESS_TOKEN_SECRET=...

Second, let's register the app with IBM Watson Workspace and get some keys!

  1. Go to the developer apps page
  2. On the left, enter the App Name and the Description of App
  3. Click on Add an outbound webhook
  4. Give the webhook a name (e.g. "listen for messages") and check the message-created webhook. This is how we'll listen to messages in a space
  5. In the callback URL, specify the URL for your app. This code assumes that the webhook listener is at https://yoururl/webhook so don't forget to add /webhook to the end of the URL (if you don't know where the app will be deployed, use a sample URL for now, like https://twitter.acme.com/webhook and you can modify that later)
  6. Click on Register app
  7. This will give you the App ID, App secret and webhook secret. You need to save these to environment variables called TWITTER_CLIENT_ID, TWITTER_CLIENT_SECRET, and TWITTER_WEBHOOK_SECRET respectively

At this point the webhook is not enabled since the system has not been able to verify it's up and running. While the webhook is probably running on Bluemix, it doesn't have the variables set

Configure your app

Now that we have the API keys, set them as environment variables. If you are using Bluemix, you can set them up following these steps:

  1. Go to the Bluemix console and click on your app
  2. Go to Runtime -> Environment Variables
  3. Define the 7 variables from above: TWITTER_CLIENT_ID, TWITTER_CLIENT_SECRET, TWITTER_WEBHOOK_SECRET, TWITTER_CONSUMER_KEY, TWITTER_CONSUMER_SECRET, TWITTER_ACCESS_TOKEN_KEY and TWITTER_ACCESS_TOKEN_SECRET, provide the right values and save them.
  4. Stop and start the app so the values take effect.

You are almost there!!!

Enable the webhook

  1. Assuming all environment variables are set (see above), go back to the IBM Watson Work Services apps page and click on the pencil icon to edit your app
  2. Check the box to enable your webhook
  3. Save your changes by clicking on "Edit app"

This now does an HTTP POST verification call to ensure your webhook is setup properly. This will also ensure that all your variables and code are good to go. If so, the webhook is now enabled and it's ready to be used! How easy was that!?!

Add the Twitter app to a space to test

We are almost there! Now we need to add the app to spaces that we want to listen for messages, and where the app will post messages to.

  1. Head out to IBM Watson Workspace and go to your favorite space
  2. Go into the space settings
  3. Click on Apps to go the apps menu
  4. Click on your app to add it to the space
  5. Type in @twitter #watsonworkspace and the app should respond with the 3 latest tweets that mention that hashtag

Have fun!!

Appendix: How to Start Twitter app locally

  1. Install Node.js 6+.
  2. Run npm install to install dependencies
  3. Run npm run build to compile the code
  4. Set the environment variables as per above.
  5. Run npm start
  6. The server will now run on localhost:3000.

You can use curl to submit a query and test it (or just deploy to Bluemix and test it there!) to simulate a call from Workspace to your webhook. Or check out the Greeter sample for further instructions on how to run locally.

What API does the app use?

The app uses the [Watson Work OAuth API] (https://workspace.ibm.com/developer/docs) to authenticate and get an OAuth token. It implements a Webhook endpoint according to the Watson Work Webhook API to listen to conversations and receive messages. Finally, it uses the [Watson Work Spaces API] (https://workspace.ibm.com/developer/docs) to send back greeting messages.

How can I contribute?

Pull requests welcome!

License and Dependencies

Licensed under Apache 2.0 (see LICENSE)

Depends on:

  • body-parser
  • express
  • request
  • twitter
  • request