#6 HTTP server

We're going to use the native http module to set up our first endpoint. This is done by creating a handler function.

A handler function takes http.ResponseWriter and http.Request arguments. The second argument is passed as a pointer. I don't know why that is. One reason could be that it allows to modify it, although I don't see why we would want to do that to an HTTP request object. A more likely explanation is that passing it by value would lead to copying the entire request object, which might be expensive. I'll update this section once I learn more.

More: "Because it's a large struct. Copying it would be expensive. So it's a pointer to a struct, which is common in Go when structs are large."

We write the response from the handler using the fmt.Fprintf method. This method takes as its first parameter an instance of http.ResponseWriter. The second argument is the response itself.

func getData(w http.ResponseWriter, req *http.Request) {
  fmt.Fprintf(w, "Hello world\n")
}

We register our handler by calling http.HandleFunc() method, passing the path and the handler itself.

http.HandleFunc("/data", getData)

Finally, we define a port to listen to:

http.ListenAndServe(":8080", nil)

We start the server with go run main.go command.

We can now visit http://localhost:8080 in our browser. The base path will return a 404 - we haven't defined a handler for this path. When we add /data, we get our response from the getData handler:

We can also make HTTP requests directly from the command line with cURL:

$ curl http://localhost:8080
404 page not found
$ curl http://localhost:8080/data
Hello world

Here's the complete code of our server:

package main

import (
  "fmt"
  "net/http"
)

func getData(w http.ResponseWriter, req *http.Request) {
  fmt.Fprintf(w, "Hello world\n")
}

func main() {
  http.HandleFunc("/data", getData)

  http.ListenAndServe(":8080", nil)
}