Elm Requests

elm-requests makes complex HTTP requests simpler:

• HTTP requests can be created bit by bit with a pipeline API.
• HTTP requests can be easily called in parallel.
• HTTP requests can be chained, so the result of one request automatically construct another request.
• Per-app configurations are stored in a global config so you don't have to repeat everything again on each request.

Examples

In Elm, a typical request will be

import Http

Http.get 
    { url = "http://deep-thought.com/api/v1/questions/ultimate"
    , expect = Http.expectJson ReceivedAnswer D.int
    }

Requests works similarly

import Request

Request.get 
    { url = "http://deep-thought.com/api/v1/questions/ultimate"
    , expect = Request.expectJson D.int
    }

Notice that the message is not part of the request. Request.get also do not return a Cmd msg, but rather a Request data object. We can execute it immediately by calling the function

Request.cmd simple ReceivedAnswer request 
    ==> Cmd

and obtain the same behavior as in Elm's native Http.get.

This little indirection has a huge payoff. We can now create request objects more easily using a pipeline interface. They can also be easily converted to tasks which allows chaining, batch execution, and more.

For instance, we can chain two http requests so our application only need to handle the final response:

getAnswer : Request Int
getAnswer = Request.get 
    { url = "http://deep-thought.com/api/v1/questions/ultimate"
    , expect = Request.expectJson D.int
    }

getQuestion : Int -> Request String
getQuestion answer = Request.get 
    { url = "http://deep-thought.com/api/v1/answers/" ++ String.fromInt
    , expect = Request.expectJson D.string
    }

Request.task simple getAnswer
    |> Task.andThen 
        (\answer -> 
            Request.task simple getQuestion
        )
    |> Task.attempt QuestionReceived