In python/javascript, I'm a big fan of using yield - https://leontrolski.github.io/recursing-with-yield.html
It just feels nice to me.
It just feels nice to me.
Typical example: say you need to connect to an api that has pagination. You call it and you get a chunk which has a number of things and a counter of remaining chunks. Normally this king of thing requires the caller to keep a lot of messy state.
with yield you can make an interface that the caller can just iterate over and it will give them one thing at a time and fetch the next chunk if the current one is exhausted and there are chunks remaining. The state of the "current chunk", "chunks remaining", index etc can all be enclosed so the caller doesn't need to know anything about them.
But it doesn't prevent the caller from passing the handle to another function, or copying the handle, which could mess things up, and I suppose `yield` helps with that.
But most of all, I'd say it makes it easier to write the callee code, not the caller code, since the callee does not need to "reify" the state of "current chunk" and "chunks remaining" into an object, nor (more crucially) the state of loops and other control flow.