Use RFC 7919 DH groups + Remove DH generation

[buchdag]

This PR brings acme-companion's DH group creation in line with nginx-proxy : nginx-proxy/nginx-proxy#1797

This PR replaces the Diffie-Hellman parameter generation with standardized RFC 7919 DH groups (2048, 3072, 4096 bit lengths). These files are available from this repo These DH groups are advised to use vs generating your own. The openssl wiki advises adopting RFC 7919 too and provides the ASCII PEM encoded content with which you can verify against the param files in this PR as the same.

Thanks to @polarathene for the original idea and work.

[polarathene]

LGTM besides a few nit-picks.

Also it might not have been visible in your editor due to your tab-width but Github renders a wider tab-width where it's rather obvious that there is mixed indentation of some lines with spaces and others with tabs, you probably want to normalize that.

[polarathene]

There's some inconsistency throughout the diff with vars.

General advice you'd get from a linter like shellcheck is to prefer "${VAR_NAME}", and within [[ ]] conditionals you can avoid the quotes IIRC (but not within the other kind [ ] that I generally avoid myself).

[buchdag]

I'm already using shellcheck 🙂

As you can see it's not complaining about "${VAR_NAME}" vs "$VAR_NAME" and require quotes on the right-hand side of == inside [[ ]]. Quoting the left-hand side is just an habit I picked up from the defensive / safe / whatever bash programming tenet of "quote all uses of variables".

I'm personally not a big fan of using "${VAR_NAME}" over "$VAR_NAME", my personal rule of thumb is to use {} only when I'm doing variable manipulation / expansion / interpolation, like described here.

I'll make at least the check_dh_group function internally consistant but clearly there are inconsistencies here and there on the whole code base. 😦

[buchdag]

Also it might not have been visible in your editor due to your tab-width but Github renders a wider tab-width where it's rather obvious that there is mixed indentation of some lines with spaces and others with tabs, you probably want to normalize that.

No, I'm aware of this, I think it's been present in the code since before I started contributing to this project and I never got to normalize the situation.