Since mutate can dynamically iterate over keys, can the function version of the data update include the source key as an argument?

[toddtarsi]

As an example, I get a big request thing back. I want to write an array of objects to the cache (normalization etc). I compose my list of paths and check if the keys match, and then in the data function, I just have no way of telling where I am if its the first write to a normalized key. If the second argument was the key we were on, I could very easily write an array of objects in one update. As it is, I have to call mutate N times, and I can see the write update split over multiple renders, which sucks.

I'd be happy to take a crack at doing this myself.

[toddtarsi]

To provide a code example of what I mean, I'd love it if this worked:

const cacheUpdates = { '/pathA': 'foo', '/pathB': 'bar' };
mutate(
  (key) => typeof key === 'string' && Object.keys(cacheUpdates).includes(key),
  (data, key) => cacheUpdates[key],
  false
);

[toddtarsi]

Then again, thinking about it, these records won't exist anyway, so they'll never get iterated on the first pass. Hmm, I'll have to think about this more. I guess I wish I could pass in an array of keys and data functions/objects, which would also make this really easy. As it is right now, I feel like the general mutate case is not quite sufficient and I would like to enhance this if possible.