gitwishes

Tweets at https://twitter.com/gitwishes.

Setup

• developed w/ python 3.6
• AWS account w/ permissions to create
  • CloudFormation stacks
  • DynamoDB tables
  • Lambda functions
  • IAM roles

Usage

1. Clone the repo, virtualenv, yada yada
2. Install the "dev" requirements: pip install -r requirements_dev.txt
   1. note: you don't need to install the Lambda function requirements (requirements.txt) unless you're testing the function locally
3. create/determine an s3 bucket to use. The packaged Lambda function and dependencies will be uploaded here
4. copy example.env to .env and populate with your settings
5. assemble the Lambda function code: invoke build-deps
6. process the CloudFormation template and upload the code: invoke package
7. deploy the CloudFormation stack: invoke deploy
8. invoke clean will clean up the locally packaged dependencies and the stuff on s3 (to save $$)

Those last three steps can be combined: invoke build-deps package deploy

You can wipe everything and start clean with: invoke clean delete build-deps package deploy

Resource details

You can see the list of created AWS resources by viewing the CloudFormation stack at https://console.aws.amazon.com/cloudformation/home

The "workflow" starts with two Cloudwatch Event Rules: one that triggers once a day, the other every four hours (would be trivial to parameterize those intervals).

The once-a-day event triggers an AWS Lambda function to search github commits for messages from the previous day matching variations of "I wish", "I really hope", etc. Results are filtered a bit, e.g. by search result score and @dariusk's wordfilter. Messages are then stored in DynamoDB with a 1-day expiration.

The every-4-hours event triggers the same Lambda function to pull off the highest scoring message from the DynamoDB table and tweet it.

There's also an IAM::Role that gets created to allow the Lambda function to execute and access DynamoDB.