Getting Started with cURL

Whenever we need to talk to the server we do it with the help of browser, where we enter the query in url and it lets you talk with the server. But how to programmers talk to servers without a browser.

When we open a website, our browser quickly sends a request to a server, waits for the response and shows us the result.

But as developers, they talk to the servers directly, without browsers, UI or buttons. All this is possible with the help of cURL.

Just like DNS and network devices, cURL is one of the tools that looks scary at first but its actually very simple once we understand why.

What is a Server?

A server is just a computer whose job is to;

Receive Request → Answer based on the request → send back response

Example:

In a shop, we ask something to the shopkeeper, the shopkeeper processes our request and then gives us a response in a product, receipt or error.

To communicate with a server, we need to send messages.

What is cURL?

cURL stands for client URL. In simple terms, cURL is a tool that lets us send requests to a server from a terminal, instead of clicking links in browser, we can type commands. We can think of it like Browsers have visual way to talk to the servers while cURL has text based way to talk to servers. It same conversation but different medium.

Why programmers Need cURL

Programmers use cURL because,

• APIs dont have buttons

• Backend services talk via HTTP

• Debugging is faster without UI

• It works everywhere on every os.

Some common real world uses are that we can Test APIs quickly, then check if a server is up, debug headers and responses and Automate backend tasks.

While designing backend cURL becomes a daily need for backend devs.

How cURL talks to a server

Making your first HTTP request using cURL

Lets start with simplest thing.

cURL example.com

What this does:

• Sends a GET request to example.com,

• Then fetches whatever the server sends back and then

• Prints it back in the terminal

We just asked the server for data where no browser was involved

Understanding Request and Response

Every HTTP conversation has 2 parts, the request and the response

1. Request

   What we send to the server, like asking a question to the server with:

   • URL

   • Method(GET/POST)

   • Optional data

2. Response

   What the server sends back like answering the given question with

   • Status code which can be 200, 404, 500

   • Data (HTML, JSON , text)

Example:

Status: 200 OK

Data <html>…</html>

Here status tells us what has happened and data tells us what we have received.

Talking to APIs using cURL

APIs are servers that usually return JSON, not HTML

A simple GET request to an API: cURL https://api.github.com

We can see the JSON output instead of a webpage HTML, That is cURL talking to an API directly

Introducing GET and POST Methods

GET: Ask for data

As the name itself tells us that GET means to GET or retrieve data.

cURL https://fake-json-api.mock.beeceptor.com/users

We can use GET when, we want to fetch data, and when we dont want to change anything in the data

POST: Send data

Post means to change or do something from our end

cURL -X POST “https://api.restful-api.dev/objects” -d ‘{ "name": "Apple MacBook Pro 16", "data": { "year": 2019, "price": 1849.99, "CPU model": "Intel Core i9", "Hard disk size": "1 TB" } }’

This adds the data into the JSON;

{
   "id": "7",
   "name": "Apple MacBook Pro 16",
   "data": {
      "year": 2019,
      "price": 1849.99,
      "CPU model": "Intel Core i9",
      "Hard disk size": "1 TB"
   },
   "createdAt": "2022-11-21T20:06:23.986Z"
}

Where cURL fits in Backend Development

Frontend (Browser) → Backend API ← cURL

cURL sits right next to Backend systems for Testing endpoints, debugging failures and reproducing bugs. If our API works in cURL then ut usually works everywhere.

When cURL Doesn’t behave properly

When something works in the browser but fails in cURL, this is very common situation.

API works in the browser but not in cURL

Most likely this happens due to some reasons:

1. Missing headers

   Browsers automatically send headers like Authorization, Content-Type, and cookies while cURL send nothing unless we tell him to.

2. Authentication required

   Browser might already be logged in but cURL is an anonymous client

3. Many APIs will return

   • 401 Unauthorized - No Access

   • 403 Forbidden - Auth required

API returns weird or empty data

Here we need to check that are we calling the correct endpoint and also check that if its a GET or POST request.

cURL command runs but shows an error

Here we need to look for HTTP status code and an error message, we should not ignore errors they usually tell us that where exactly we are wrong.

Common mistakes beginners make with cURL

1. Expecting a UI where cURL shows raw data not pretty pages

2. Forgetting the URL scheme we need to add https before;

   Example:

   • cURL example.com is wrong way

   • cURL https://example.com is correct

3. We should not panic when output looks very messy because API speaks in JSON and not english

Conclusion

cURL is not a scary networking tool, its simply A direct conversation between you and the server. Once we get comfortable with sending requests and reading responses then we unlock new potential of API debbugging , Backend confidence and troubleshoot an error.