sutaba-server

Twitter bot that recognize whether the replied image is photographed in Starbucks or not.

Configuration

You can specify server configurations by flags or config file. Also, you must specify twitter credentials via environment variable.

flags

Start server

Usage:
  sutaba-server start [flags]

Flags:
      --bot-id int                 bot twitter user id
      --classifier-server string   classifier server url
      --error-message string       text of tweet for error notification
  -h, --help                       help for start
      --keyword string             process only tweets which contain this value
      --owner-id int               owner twitter user id (error tweet will be send to owner if something is failed)
      --sorry-message string       text of tweet to send to user if process is failed

Global Flags:
      --config string   config file (default is $HOME/.sutaba-server.yaml)

config file

See config.yml

Twitter credentials

You must specify below environment variables.

• TWITTER_CONSUMER_KEY
• TWITTER_CONSUMER_SECRET
• OWNER_TWITTER_ACCESS_TOKEN
• OWNER_TWITTER_ACCESS_TOKEN_SECRET
• BOT_TWITTER_ACCESS_TOKEN
• BOG_TWITTER_ACCESS_TOKEN_SECRET

Run server

binary

Download from GitHub Releases and put it anywhere in your executable path.

$ sutaba-server start --config ./config.yml

from source

$ git clone https://github.com/mpppk/sutaba-server
$ cd sutaba-server
$ go run main.go start --config ./config.yml

Docker

$ git clone https://github.com/mpppk/sutaba-server
$ cd sutaba-server
$ docker-compose up

Deploy

This repository includes deploy scripts for GCP(Container Registry and Cloud Run).
Before run scripts, update scripts by your project id.

$ ./scripts/gcp/submit_container_image.sh
$ ./scripts/gcp/deploy_to_cloud_run.sh