KiCad file render service

Render KiCad schematic and layout files into browser-viewable formats on demand.

Features

• Downloads KiCad schematics and layouts from Bitbucket and returns a rendered SVG image or PDF document on demand.
• Bitbucket Fileviewer to view KiCad schematics and layouts in Bitbucket.

Installation and running

Using Docker compose

TODO: use and document this

Using Docker

• Start a docker network:

docker network create kicad-file-render-network

• Start the connections_db Redis database connected to the network:

docker run --name connections-db -d --network kicad-file-render-network redis

Hint: You can use redis:alpine if you want a smaller container size Another hint: pass -p 6379:6379 to allow you host to connect to the DB for debugging purposes.

• Create a secret key:

This service requires a (secret) key in order to safely store secrets. You can pass it:

As an environmental variable:

export KICAD_FILE_RENDERER_KEY=$(dd if=/dev/random bs=3 count=16 2>/dev/null | base64)

Or as a Docker secret (requires a Docker swarm):

dd if=/dev/random bs=3 count=16 2>/dev/null | base64 | docker secret create file-render-key-test -

• Run the service

The service needs to know where it's running so it can tell Bitbucket and other services. You can pass it using the ADDRESS environmetal variable.

It also needs the secret created in the previous step.

This can also be passed as an environmental variable:

docker run -p 5000:5000 --name kicad-file-render-service -d --network kicad-file-render-network -e KICAD_FILE_RENDERER_KEY=$(KICAD_FILE_RENDERER_KEY) productize/kicad-file-render-service

Or by using the docker secret

Note: If you don't want to use Docker hub, you can also build the docker image yourself from this repo using:

docker build .

Yet another note: For debugging purposes it's useful to run the docker image interactively using the source files on the host. This can be done with Docker volumes:

docker run -p 5000:5000 --name kicad-file-render-service -it --network kicad-file-render-network -v `pwd`:/kicad-file-render-service/ productize/kicad-automation-scripts

Manually

• Install the dependencies, e.g:

apt-get install -y kicad python3 python3-pip xvfb recordmydesktop xdotool
pip3 install kicad_automation_scripts/eeschema/requirements.txt

See the Dockerfiles for an up-to date list of dependencies.

• Export the (secret) KICAD_FILE_RENDERER_KEY variable, e.g:

export KICAD_FILE_RENDERER_KEY=$(dd if=/dev/random bs=3 count=16 2>/dev/null | base64)

• Run the flask app:

export FLASK_APP=bitbucket.py
ADDRESS="https://kicad-file-render-service.your-domain.com" python3 -m flask run

Set up an HTTPS proxy to port 5000

Bitbucket requires the service to run on HTTP with a verifiable, valid certificate. If you don't have one you could give letsencrypt a try.

The easiest way to achieve this is setting up an HTTPS proxy towards the port this app is configured to use (5000 by default / in example) in your favorite HTTPS server. E.g on NGINX:

server {
    listen      443 default;
    server_name kicad-file-renderer.productize.be;

    ssl on;
    ssl_certificate     /your/certs/chained-certificate.crt;
    ssl_certificate_key /your/certs/certificate-key.key;
    location / {
        proxy_pass http://localhost:5000;
    }
}

Wishlist

• Add support for KiCad schematic libraries and layout footprints
• Add support for files hosted on Github and Gitlab (no way to add file viewers in Github, but could be done by e.g embedding in e.g README's).
• Create a diffview for reviewing changesets
  • Incorporate the diffview in Pullrequest views (not supported in Bitbucket).
• A home page
• Run it as a service