Lorna Mitchell

Vonage Team Member

Lorna is a software engineer with an incurable blogging habit. She tries to tame words and code in equal measure.

[Read more]
< Tutorial />

Download Vonage Voice API Recordings with Golang

Last updated on May 05, 2021

We love the Voice API and all the fun things you can do with it. Today I'm going to show you a Golang application I'm using to download the recordings of the calls.

Vonage API Account

To complete this tutorial, you will need a Vonage API account. If you don’t have one already, you can sign up today and start building with free credit. Once you have an account, you can find your API Key and API Secret at the top of the Vonage API Dashboard.

This tutorial also uses a virtual phone number. To purchase one, go to Numbers > Buy Numbers and search for one that meets your needs.

Screenshot of new Meetings API session in progress
Start developing in minutes with free credits on us. No credit card required!

When I make a call and record it, the API sends a webhook at the end of the call with all the information about the recording. My application receives this webhook and downloads the recording file itself.

Configure the Recording URL

When I add a record action to a Voice API NCCO, I can set the eventUrl to receive the recording notification. This incoming webhook arrives when the call is finished, and includes information about the recording and a link to download it.

My record action looks like this:

{
  "action": "record",
  "eventUrl": ["https://76b239af.ngrok.io/recording"]
}

There are two things to notice here:

  1. This eventUrl is actually an array (it catches me out about one time in every three!)
  2. I'm using Ngrok to provide a publicly-available URL to my local development platform. You can read about Ngrok on the Developer Portal if you'd like to learn more about using this approach.

Receive the Webhook

The incoming webhook is JSON-formatted and looks something like this:

{
    "start_time": "2020-05-06T13:34:21Z",
    "recording_url": "https://api.nexmo.com/v1/files/6d29bd8d-e6ff-45b9-9379-2843fe7b37fe",
    "size": 15822,
    "recording_uuid": "692100cb-e4ef-4f18-ab90-2a09573aecb5",
    "end_time": "2020-05-06T13:34:25Z",
    "conversation_uuid": "CON-55970ffd-a6b7-4d18-b3b6-088c03ea49f1",
    "timestamp": "2020-05-06T13:34:25.771Z"
}

In my Golang code, I'm going to handle an incoming request to /recording and download the file, saving it to disk.

To authenticate, you will need a (JWT)[https://developer.nexmo.com/concepts/guides/authentication#json-web-tokens-jwt) which you can generate programmatically or from the command line. This example expects it to be in an environment variable JWT.

package main

import (
    "encoding/json"
    "fmt"
    "io"
    "net/http"
    "os"
)

type RecordingWebhook struct {
    StartTime        string  `json:"start_time"`
    RecordingURL     string  `json:"recording_url"`
    Size             float64 `json:"size"`
    RecordingUUID    string  `json:"recording_uuid"`
    EndTime          string  `json:"end_time"`
    ConversationUUID string  `json:"conversation_uuid"`
    Timestamp        string  `json:"timestamp"`
}

func downloadRecording(w http.ResponseWriter, r *http.Request) {
    jwt := os.Getenv("JWT")

    // Get data from incoming webhook
    data := RecordingWebhook{}
    err := json.NewDecoder(r.Body).Decode(&data)
    if err != nil {
        http.Error(w, err.Error(), http.StatusBadRequest)
        return
    }

    fmt.Println("Recording URL: " + data.RecordingURL)
}

func main() {
    http.HandleFunc("/recording", downloadRecording)
    if err := http.ListenAndServe(":8080", nil); err != nil {
        panic(err)
    }
}

The entry point here is the main() function at the end of the code sample. It registers a /recording route and then starts a web server running on port 8080.

When I first run the code with go run main.go, not much happens! That's because the web server is running and waiting for a request to arrive. When it does, if the route matches then it calls the downloadRecording() function and the interesting stuff starts!

This code first parses the incoming data; it is a POST request with a JSON body so I defined a struct that I could decode the data into. Once we have the URL, the program outputs it.

Download and Save the Recording

If things went well to this point then we can move on and add the steps to download the recording and save the file. When complete, the full downloadRecording() function looks like this:

func downloadRecording(w http.ResponseWriter, r *http.Request) {
    jwt := os.Getenv("JWT")

    // Get data from incoming webhook
    data := RecordingWebhook{}
    err := json.NewDecoder(r.Body).Decode(&data)
    if err != nil {
        http.Error(w, err.Error(), http.StatusBadRequest)
        return
    }

    fmt.Println("Recording URL: " + data.RecordingURL)
    // prepare and download the recording, with auth
    req, err := http.NewRequest("GET", data.RecordingURL, nil)
    if err != nil {
        panic(err)
    }

    req.Header.Set("Authorization", "Bearer "+jwt)
    client := &http.Client{}
    resp, err := client.Do(req)
    if err != nil {
        panic(err)
    }

    defer resp.Body.Close()

    // now write to a local file
    filename := data.RecordingUUID + ".mp3"
    out, err := os.Create(filename)
    if err != nil {
        panic(err)
    }
    defer out.Close()

    _, fileErr := io.Copy(out, resp.Body)
    if fileErr != nil {
        panic(fileErr)
    }

    // Good! acknowledge it
    w.Write([]byte("OK"))
}

Downloading recordings requires credentials, so this code shows how to add a JWT to the request before sending it.

Finally, the response to the download request is written to a local file. I used the recording ID as the file name since I know it will be unique; if it makes more sense to name with timestamps or something else then you could definitely do that in your own application.

Handling Recordings in your own Applications

Today's example is basic but does show some key ingredients of working with Voice API; receiving webhooks in response to events and sending credentials to download the recordings. If you're using something similar in your own applications or build on this example, let us know! We always enjoy hearing what you are all working on.