Quick Start

Open a Gun connection. Wait until it is ready. Send a GET request. Collect the full response in memory.

import gleam/int
import gleam/io
import gleam/result
import gluegun/client
import gluegun/connection
import gluegun/request
import gluegun/response

pub fn main() {
  let timeout = connection.Milliseconds(5000)

  use conn <- result.try(
    connection.options()
    |> connection.open(host: "example.com", port: 80),
  )
  use _protocol <- result.try(connection.await_up(conn, timeout))

  use res <- result.try(
    client.new(request.Get, "/")
    |> client.with_timeout(timeout: timeout)
    |> client.send(connection: conn),
  )

  io.println("status: " <> int.to_string(response.status(res)))

  case response.body_text(res) {
    Ok(text) -> io.println(text)
    Error(_) -> io.println("response body was not UTF-8")
  }

  connection.close(conn)
}

Key idea

Gluegun keeps connection setup and requests separate:

1. Open a connection to a host and port.
2. Wait until Gun reports the negotiated protocol.
3. Send requests with paths such as /, /api/items, or /health.
4. Close or shut down the connection when you are finished.

If you have a full URL, parse it with gleam/uri before you call Gluegun. Gluegun accepts only the separate host, port, transport, path, and query values.

connection.Milliseconds(Int) makes a finite Timeout. Use connection.Infinity to wait without a limit. The same Timeout value applies to connection readiness, request bodies, and message receives.

Use connection.close for usual teardown. Use connection.shutdown only when a connection seems stuck. shutdown stops the Gun process immediately, without a graceful close.

For full module, type, and function details, see the API reference.