Add a core utility to strip magic keys and other keys from a component

[dead-claudia]

Mithril version:

Browser and OS:

Project:

Is this something you're interested in implementing yourself? Yes

Description

Continuation of #1986

Specifically, I'm suggesting bringing in this utility into core. Already just got bit writing a built-in utility, and I know not everyone reads the docs closely. Plus, having a core utility that contradicts this doesn't really make sense, and I don't expect everyone to read the docs closely enough to pick this up. So instead, we should just provide a built-in utility to strip all that crap out for people.

Why

Well, if you want to know how easy it is to get bit without realizing it, take a look at #2471.

Possible Implementation

https://github.com/isiahmeadows/mithril-helpers/blob/master/censor.js with the alterations needed to fuse into core. (It's 324 bytes min+gzip standalone.)

Open Questions

[dead-claudia]

And to append to that, the docs don't make it clear it applies even in cases like this, the moral equivalent of what I did:

const {component, options, href, ...rest} = vnode.attrs

return m(component, {
    ...rest,
    href: prefix + href,
    onclick() { ... }
}, vnode.children)

[dead-claudia]

BTW, I don't plan to "fix" m.route.Link to follow this, but instead plan to add a utility to get out of the way of users. It won't likely hit v2.0.0, but my hope is that it hits v2.1.0.

/cc @barneycarroll @StephanHoyer if you all have thoughts.

[dead-claudia]

Also, I don't plan to implement all of this, just implementing m.censor. It's relatively small, somewhat surprisingly so for being as optimized as it is.

[dead-claudia]

I just did some measurements, and it'd increase the min+gzip bundle size from 9546 bytes to 9673 bytes, an increase of just under 130 bytes. So it's relatively small.