# Make HTTP Requests with Node.js
HTTP requests are fundamental to working with APIs or other web services. You can make HTTP requests to retrieve data from APIs, fetch HTML from websites, or do pretty much anything your web browser can do.
Below, we'll review how to make HTTP requests using Node.js code on Pipedream.
We'll use the axios
(opens new window) and got
(opens new window) HTTP clients in the examples below, but you can use any npm package you'd like on Pipedream, so feel free to experiment with other clients, too.
If you're developing Pipedream components, you may find the @pipedream/platform
version of axios
helpful for displaying error data clearly in the Pipedream UI.
If you're new to HTTP, see our glossary of HTTP terms (opens new window) for a helpful introduction.
- Basic axios usage notes
- Send a GET request to fetch data
- Send a POST request to submit data
- Pass query string parameters to a GET request
- Send a request with HTTP headers
- Send a request with a secret or API key
- Send multiple HTTP requests in sequence
- Send multiple HTTP requests in parallel
- Send a multipart/form-data request
- Download a file to the /tmp directory
- Upload a file from the /tmp directory
- IP addresses for HTTP requests made from Pipedream workflows
- Use an HTTP proxy to proxy requests through another host
- Stream a downloaded file directly to another URL
- Catch and process HTTP errors
- Paginating API requests
- Send GraphQL request
# Basic axios
usage notes
To use axios
on Pipedream, you'll just need to import the axios
npm package:
import axios from "axios";
You make HTTP requests by passing a JavaScript object (opens new window) to axios
that defines the parameters of the request. For example, you'll typically want to define the HTTP method and the URL you're sending data to:
{
method: "GET",
url: `https://swapi.dev/api/films/`
}
axios
returns a Promise (opens new window), which is just a fancy way of saying that it makes the HTTP request in the background (asynchronously) while the rest of your code runs. On Pipedream, all asynchronous code must be run synchronously, which means you'll need to wait for the HTTP request to finish before moving on to the next step. You do this by adding an await
in front of the call to axios
.
Putting all of this together, here's how to make a basic HTTP request on Pipedream:
const resp = await axios({
method: "GET",
url: `https://swapi.dev/api/films/`,
});
The response object resp
contains a lot of information about the response: its data, headers, and more. Typically, you just care about the data, which you can access in the data
property of the response:
const resp = await axios({
method: "GET",
url: `https://swapi.dev/api/films/`,
});
// HTTP response data is in the data property
const data = resp.data;
Alternatively, you can access the data using object destructuring (opens new window), which is equivalent to the above and preferred in modern JavaScript:
const { data } = resp;
# Send a GET
request to fetch data
Make a request to retrieve Star Wars films from the Star Wars API:
Copy this workflow to run this example on Pipedream (opens new window).
# Send a POST
request to submit data
POST sample JSON to JSONPlaceholder (opens new window), a free mock API service:
When you make a POST
request, you pass POST
as the method
, and include the data you'd like to send in the data
object.
Copy this workflow to run this example on Pipedream (opens new window).
# Pass query string parameters to a GET
request
Retrieve fake comment data on a specific post using JSONPlaceholder (opens new window), a free mock API service. Here, you fetch data from the /comments
resource, retrieving data for a specific post by query string parameter: /comments?postId=1
.
You should pass query string parameters using the params
object, like above. When you do, axios
automatically URL-encodes (opens new window) the parameters for you, which you'd otherwise have to do manually.
Copy this workflow to run this code on Pipedream (opens new window).
# Send a request with HTTP headers
You pass HTTP headers in the headers
object of the axios
request:
import axios from "axios";
// Make an HTTP POST request using axios
const resp = await axios({
method: "POST",
url: `https://jsonplaceholder.typicode.com/posts`,
headers: {
"Content-Type": "application/json",
},
data: {
name: "Luke",
},
});
# Send a request with a secret or API key
Most APIs require you authenticate HTTP requests with an API key or other token. Please review the docs for your service to understand how they accept this data.
Here's an example showing an API key passed in an HTTP header:
import axios from "axios";
// Make an HTTP POST request using axios
const resp = await axios({
method: "POST",
url: `https://jsonplaceholder.typicode.com/posts`,
headers: {
"Content-Type": "application/json",
"X-API-Key": "123", // API KEY
},
data: {
name: "Luke",
},
});
Copy this workflow to run this code on Pipedream (opens new window).
# Send multiple HTTP requests in sequence
There are many ways to make multiple HTTP requests. This code shows you a simple example that sends the numbers 1
, 2
, and 3
in the body of an HTTP POST request:
import axios from "axios";
export default defineComponent({
async run({ steps, $ }) {
// We'll store each response and return them in this array
const responses = [];
for await (const num of [1, 2, 3]) {
const resp = await axios({
method: "POST",
url: "https://example.com",
data: {
num, // Will send the current value of num in the loop
},
});
responses.push(resp.data);
}
return responses;
},
});
This sends each HTTP request in sequence, one after another, and returns an array of response data returned from the URL to which you send the POST request. If you need to make requests in parallel, see these docs.
Copy this workflow (opens new window) and fill in your destination URL to see how this works. This workflow iterates over the value of a Pipedream step export - data returned from a previous step. Since you often want to iterate over data returned from a Pipedream action or other code step, this is a common use case.
# Send multiple HTTP requests in parallel
Sometimes you'll want to make multiple HTTP requests in parallel. If one request doesn't depend on the results of another, this is a nice way to save processing time in a workflow. It can significantly cut down on the time you spend waiting for one request to finish, and for the next to begin.
To make requests in parallel, you can use two techniques. By default, we recommend using promise.allSettled
(opens new window), which makes all HTTP requests and returns data on their success / failure. If an HTTP request fails, all other requests will proceed.
import axios from "axios";
export default defineComponent({
async run({ steps, $ }) {
const arr = [
"https://www.example.com",
"https://www.cnn.com",
"https://www.espn.com",
];
const promises = arr.map((url) => axios.get(url));
return Promise.allSettled(promises);
},
});
First, we generate an array of axios.get
requests (which are all Promises (opens new window)), and then call Promise.allSettled
to run them in parallel.
When you want to stop future requests when one of the requests fails, you can use Promise.all
(opens new window), instead:
import axios from "axios";
export default defineComponent({
async run({ steps, $ }) {
const arr = [
"https://www.example.com",
"https://www.cnn.com",
"https://www.espn.com",
];
const promises = arr.map((url) => axios.get(url));
return Promise.all(promises);
},
});
The Mozilla docs expand on the difference between these methods, and when you may want to use one or the other:
The
Promise.allSettled()
method returns a promise that resolves after all of the given promises have either fulfilled or rejected, with an array of objects that each describes the outcome of each promise.
It is typically used when you have multiple asynchronous tasks that are not dependent on one another to complete successfully, or you'd always like to know the result of each promise.
In comparison, the Promise returned byPromise.all()
may be more appropriate if the tasks are dependent on each other / if you'd like to immediately reject upon any of them rejecting.
# Send a multipart/form-data
request
Copy this workflow (opens new window) to run this example.
# Download a file to the /tmp
directory
This example shows you how to download a file to a file in the /tmp
directory. This can be especially helpful for downloading large files: it streams the file to disk, minimizing the memory the workflow uses when downloading the file.
import { pipeline } from "stream/promises";
import fs from "fs";
import got from "got";
export default defineComponent({
async run({ steps, $ }) {
// Download the webpage HTML file to /tmp
return await pipeline(
got.stream("https://example.com"),
fs.createWriteStream('/tmp/file.html')
);
}
})
Copy this workflow (opens new window) to run this example.
# Upload a file from the /tmp
directory
This example shows you how to make a multipart/form-data
request with a file as a form part. You can store and read any files from the /tmp
directory.
This can be especially helpful for uploading large files: it streams the file from disk, minimizing the memory the workflow uses when uploading the file.
import axios from "axios";
import fs from "fs";
import FormData from "form-data";
export default defineComponent({
async run({ steps, $ }) {
const formData = new FormData();
formData.append("file", fs.createReadStream('/tmp/file.pdf'));
const headers = formData.getHeaders();
const config = {
method: "POST",
url: "https://example.com",
headers,
data: formData,
};
return await axios(config);
}
});
Copy this workflow (opens new window) to run this example.
# IP addresses for HTTP requests made from Pipedream workflows
By default, HTTP requests made from Pipedream can come from a large range of IP addresses. If you need to restrict the IP addresses HTTP requests come from, you have two options:
- Use a Pipedream VPC to route all outbound HTTP requests through a single IP address
- If you don't need to access the HTTP response data, you can use
$send.http()
to send requests from a limited set of IP addresses.
# Use an HTTP proxy to proxy requests through another host
By default, HTTP requests made from Pipedream can come from a range of IP addresses. If you need to make requests from a single IP address, you can route traffic through an HTTP proxy:
import axios from "axios";
import httpsProxyAgent from "https-proxy-agent";
export default defineComponent({
props: {
user: {
type: 'string',
label: 'Username',
description: 'The username for the HTTP proxy authentication',
},
pass: {
type: 'string',
label: 'Password',
secret: true,
description: 'The password for the HTTP proxy authentication',
},
host: {
type: 'string',
label: "HTTP Proxy Host",
description: "The URL for the HTTP proxy",
},
port: {
type: "string",
label: "Port",
description: "The port the HTTP proxy is accepting requests at",
},
target_host: {
type: 'string',
label: "Target Host",
description: "The URL for the end target to reach through the proxy",
},
method: {
type: 'string',
default: 'GET',
label: "HTTP method",
description: "The HTTP method to use to reach the end target host"
},
body: {
type: 'object',
label: "HTTP body",
description: "The HTTP body payload to send to the end target host"
}
},
async run({ steps, $ }) {
const { user, pass, host, port, target_host, method } = this;
const agent = new httpsProxyAgent(`http://${user}:${pass}@${host}:${port}`);
const config = {
method,
url: target_host,
body,
httpsAgent: agent,
};
return await axios.request(config);
}
});
Copy this workflow to run this code on Pipedream (opens new window).
# Stream a downloaded file directly to another URL
Sometimes you need to upload a downloaded file directly to another service, without processing the downloaded file. You could download the file and then upload it to the other URL, but these intermediate steps are unnecessary: you can just stream the download to the other service directly, without saving the file to disk.
This method is especially effective for large files that exceed the limits of the /tmp
directory.
Copy this workflow (opens new window) or paste this code into a new Node.js code step:
import stream from "stream";
import { promisify } from "util";
import got from "got";
export default defineComponent({
async run({ steps, $ }) {
const pipeline = promisify(stream.pipeline);
await pipeline(
got.stream("https://example.com"),
got.stream.post("https://example2.com")
);
}
});
You'll want to replace https://example.com
with the URL you'd like to stream data from, and replace https://example2.com
with the URL you'd like to send the data to. got
streams the content directly, downloading the file using a GET
request and uploading it as a POST
request.
If you need to modify this behavior, see the got
Stream API (opens new window).
# Catch and process HTTP errors
By default, axios
throws an error when the HTTP response code is in the 400-500 range (a client or server error). If you'd like to process the error data instead of throwing an error, you can pass a custom function to the validateStatus
property:
import axios from "axios";
export default defineComponent({
async run({ steps, $ }) {
const resp = await axios({
url: "https://httpstat.us/400",
validateStatus: () => true, // will not throw error when axios gets a 400+ status code (the default behavior)
});
if (resp.status >= 400) {
this.debug = resp;
throw new Error(JSON.stringify(resp.data)); // This can be modified to throw any error you'd like
}
return resp;
}
});
See the axios
docs (opens new window) for more details.
# Paginating API requests
When you fetch data from an API, the API may return records in "pages". For example, if you're trying to fetch a list of 1,000 records, the API might return those in groups of 100 items.
Different APIs paginate data in different ways. You'll need to consult the docs of your API provider to see how they suggest you paginate through records.
# Send GraphQL request
Make a GraphQL request using the graphql-request
NPM package:
import { graphql } from 'graphql'
import { request, gql } from 'graphql-request'
export default defineComponent({
async run({ steps, $ }) {
const document = gql`
query samplePokeAPIquery {
generations: pokemon_v2_generation {
name
pokemon_species: pokemon_v2_pokemonspecies_aggregate {
aggregate {
count
}
}
}
}
`
return await request('https://beta.pokeapi.co/graphql/v1beta', document)
},
})
The graphql package is required
The graphql
package is required for popular GraphQL clients to function, like graphql-request
and urql
.
Even though you will not need to use the graphql
code itself in your code step, it's required to import it in order for graphql-request
to function.
# Send an authenticated GraphQL request
Authenticate your connected accounts in Pipedream with GraphQL requests using the app
prop:
import { graphql } from 'graphql'
import { GraphQLClient, gql } from 'graphql-request'
export default defineComponent({
props: {
github: {
type: 'app',
app: 'github'
}
},
async run({ steps, $ }) {
const me = gql`
query {
viewer {
login
}
}
`
const client = new GraphQLClient('https://api.github.com/graphql', {
headers: {
authorization: `Bearer ${this.github.$auth.oauth_access_token}`,
},
})
return await client.request(me)
},
})
Alternatively, you can use Environment Variables as well for simple API key based GraphQL APIs.