For a uM rule like:
foo.com bar.com css allow
which means "allow foo.com to fetch css from bar.com", the corresponding uBO static rule is:
@@||bar.com^$domain=foo.com,css,allow
The full list of things that can be allow/block'd by uBO is at
https://github.com/gorhill/uBlock/wiki/Static-filter-syntax#...I have a "block everything by default" rule at the top that's:
*$css,font,frame,media,object,ping,script,websocket,xhr
*$image,redirect=1x1.gif
*$csp=worker-src 'none'
@@*$1p,css,frame,image
which means:
1. Block a bunch of things by default.
2. Block images by replacing them with the built-in 1x1 GIF instead of canceling the request.
3. Disable web workers by setting the CSP worker-src.
4. Override the previous rules by allowing first-party CSS, frames and images. (The @@ means it's an override rule.)
(The fact that my default is to block everything is why the first example I gave above starts with @@ too.)
Web workers can be allowed on a per-site basis by overriding the csp directive with a reset:
@@||foo.com^$csp
Lastly, I have a dynamic rule to allow `` tags to be rendered:
no-scripting: * true
Then, for every static rule where I enable JS for a domain, I add a corresponding `no-scripting: $domain false` in the dynamic rules.
It's annoying to have to move between static and dynamic rules when deciding to enable JS on a site, but I'm not sure there's a better way. Neither static nor dynamic rules individually support everything that uM could do - static rules can't block inline JS nor render `` content, and dynamic rules can't block every kind of request.
Static rules are also nice in that you can have empty lines and comments and arbitrary ordering of your rules, so it's easier to group rules in sections based on the domain names, add comments, etc. Dynamic rules however are like uM's rules and are forced to be sorted by domain name with no empty lines or comments.