yarn

with env variables

• CALLBACK_HOST — Internal API that called after screenshoting. default: http://localhost:9002
• POOL_SIZE — Count of tabs in browser. default require('os').cpus().length / 2
• QUEUE_NAME — Queue name in Rabbit. default howtocards:render
• RABBIT_HOST — RabbitMQ server. default amqp://localhost:5672
• RENDER_HOST — Howtocards Frontend instance. default https://howtocards.io
• UPLOADER_HOST — Image Uploader internal API. default http://localhost:4000

• yarn start to run worker
• CTRL+C — shutdown worker

docker run -v `pwd`/the-files:/files -it -e "VOLUME=/files" --expose 4000  howtocards/image-uploader/image-uploader

Worker subscribes to tasks from QUEUE_NAME at RABBIT_HOST.

Task types:

• { "user": "@sergeysova", "extra": {}, "callback": "/preview/user/@sergeysova" } — to render /@sergeysova
• { "card": "2", "extra": {}, "callback": "/preview/card/2" } — to render /open/2

Steps:

1. Task received

2. Tab consumed from pool

3. Task rendered by specified type

   • user just makes screenshot of user page
   • card makes screenshot of card, and get snapshot. Frontend should render with Slate

4. Tab returned to pool

5. Screenshot is uploaded to UPLOADER_HOST

6. If .callback present send POST request to {CALLBACK_HOST}/{callback} with JSON body { "screenshot": "/abcd/efgh/asdasd.png", "snapshot": "<div></div>" }

   • snapshot generated only for card
   • user sends only {"screenshot": ""}
   • screenshot is a path relative to https://howtocards.io/images