behaviour of ractive.toggle('foo.*.bar')

[Rich-Harris]

This SO question raises an interesting point: should ractive.toggle(), if used with a keypath pattern like foo.*.bar, toggle all the keypaths that match that pattern?

At present, it's just a simple wrapper around ractive.set():

// this...
ractive.toggle( 'foo.*.bar' );

// ...is equivalent to this...
ractive.set( 'foo.*.bar', !this.get( 'foo.*.bar' ) );

// ...which is equivalent to this. Not very helpful
ractive.set( 'foo.*.bar', true );

Possibly a niche use case but it does break expectations. The ractive.add() and ractive.subtract() convenience methods are in the same boat, though that's even more niche.

[Madgvox]

I think it makes the most sense (and I think this how I thought of it in my mind) for wildcard keypaths to act like a forEach would.

console.log( ractive.get( 'foo' ) ); // -> [ { bar: true }, { bar: false }, { bar: true } ]

ractive.toggle( 'foo.*.bar' );

console.log( ractive.get( 'foo' ) ); // -> [ { bar: false }, { bar: true }, { bar: false } ]

[fskreuz]

I think it should, if one assumes that foo is an array and * means indexes.

Wild question: What happens if foo was an object, and * means properties of an object? Should it set bar inside them? One could probably defer this as an end-user responsibility, but then...