Because currently the --print-to-pdf CLI switch does not allow disabling of
header/footer the approach uses a nodejs script which connects to chrome via
remote interface. It also executes 5s of javascript (virtual time) to ensure
that the page is fully rendered.
Check out the code by my colleague: chrome-headless-render.
You need to mount a container workdir locally to be able to get the output file.
Also because chrome uses some kernel feature for sandboxing you need to run the container
in --privileged mode. It's possible to work around this by disabling sandboxing
but this seems to work best.
The workdir is /tmp/html-to-pdf, thus running:
docker run -v /your-local-dir:/tmp/html-to-pdf --privileged pink33n/html-to-pdf --url http://google.com --pdf out.pdf
Will produce /your-local-dir/out.pdf file.
As it is an unstable version things will probably change & break, you will want your pdf output stable.
You'll want a specific version of chrome and this is the way to guarantee that.
The container versioning scheme is ${MAJOR_CHROME_VERSION}.${CONTAINER_VERSION}