When requesting from a server with JavaScript fetch API, you have to do something like
fetch(API)
.then(response => response.json())
.catch(err => console.log(err))
Here, response.json()
is resolving its promise.
The thing is that if you want to catch 404
's errors, you have to resolve the response promise and then reject the fetch promise, because you'll only end in catch
if there's been a network error. So the fetch call becomes something like
fetch(API)
.then(response => response.ok ? response.json() : response.json().then(err => Promise.reject(err)))
.catch(err => console.log(err))
This is something much harder to read and reason about. So my question is: why is this needed? What's the point of having a promise as a response value? Are there any better ways to handle this?
If your question is "why does response.json()
return a promise?" then @Bergi provides the clue in comments: "it waits for the body to load".
If your question is "why isn't response.json
an attribute?", then that would have required fetch
to delay returning its response until the body had loaded, which might be OK for some, but not everyone.
This polyfill should get you what you want:
var fetchOk = api => fetch(api)
.then(res => res.ok ? res : res.json().then(err => Promise.reject(err)));
then you can do:
fetchOk(API)
.then(response => response.json())
.catch(err => console.log(err));
The reverse cannot be polyfilled.