In this simple example of an http server in Node.js print out some of the values from the request object which is an instance of the http.ClientRequest class.

examples/node/http_client_request.js

var http = require('http');

var port = 8081;

var s = http.createServer();
s.on('request', function(request, response) {
    response.writeHead(200);
    console.log(request.method);
    console.log(request.headers);
    console.log(request.url);
    response.write('hi');
    response.end();
});

s.listen(port);
console.log('Browse to http://127.0.0.1:' + port);

I ran the above script using node examples/node/http_client_request.js it printed Browse to http://127.0.0.1:8081 so I browsed there with my regular browser.

This is what was printed on the console:

GET
{ host: '127.0.0.1:8081',
  connection: 'keep-alive',
  'cache-control': 'max-age=0',
  accept: 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
  'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.94 Safari/537.36',
  'accept-encoding': 'gzip, deflate, sdch',
  'accept-language': 'en-US,en;q=0.8,he;q=0.6,ru;q=0.4' }
/
GET
{ host: '127.0.0.1:8081',
  connection: 'keep-alive',
  accept: '*/*',
  'user-agent': 'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_2) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.94 Safari/537.36',
  'accept-encoding': 'gzip, deflate, sdch',
  'accept-language': 'en-US,en;q=0.8,he;q=0.6,ru;q=0.4' }
/favicon.ico

At first it confused me a bit, why do I have two GET request, but then I remembered, and looking more closely I can also see it. The first request was indeed to /, (as you can above the second GET line), but then another request was sent by my browser to the /favicon.ico. This is an automatic request my browser send in the hope that it will be able to put this little image on the tab where I opened the page.

I don't want this extra confusion in my research, so I switched to the curl command that is available on Linux/Unix.

Let's start again:

I ran the server:

$ node examples/node/http_client_request.js 
Browse to http://127.0.0.1:8081

and opened another shell window where I sent my request:

$ curl http://127.0.0.1:8081/

Then switched back to the original console to see the response:

GET
{ 'user-agent': 'curl/7.37.1',
  host: '127.0.0.1:8081',
  accept: '*/*' }
/

From this point I was switching back-and-forth between the two consoles.

The second request included a path on the server and a parameter with a value:

$ curl http://127.0.0.1:8081/some/path?field=value

The printout on the other console was similar to the previous printout, except of the last line that shows the url.

GET
{ 'user-agent': 'curl/7.37.1',
  host: '127.0.0.1:8081',
  accept: '*/*' }
/some/path?field=value

The last attempt was to send in a POST request using curl with some data:

$ curl --data "field=value" http://127.0.0.1:8081/

the output looked like this:

POST
{ 'user-agent': 'curl/7.37.1',
  host: '127.0.0.1:8081',
  accept: '*/*',
  'content-length': '11',
  'content-type': 'application/x-www-form-urlencoded' }
/

The first line shows that it was indeed a POST request, the header had some extra fields, but the data itself was not included.

Of course, the data needs to be read in and processed in another way.

Accepting HTTP POST request in Node.js

examples/node/http_client_request_post.js

var http = require('http');

var port = 8081;

var s = http.createServer();
s.on('request', function(request, response) {
    response.writeHead(200);
    console.log(request.method);
    console.log(request.headers);
    console.log(request.url);

    var data = '';
    request.on('data', function(chunk) {
        data += chunk.toString();
    });
    request.on('end', function() {
        console.log(data);
        response.write('hi');
        response.end();
    });

});

s.listen(port);
console.log('Browse to http://127.0.0.1:' + port);

This is another area where the non-blocking nature of Node.js is seen. Instead of just reading the data from the request object we add a callback method to the data event of the request object. This will be called every time another piece of data arrives. Of course, if all the data is only 11 characters as in our case, then this is not very interesting, but if you are sending a lot of data then it is important to read that in without blocking the rest of the site.

Now that we have a call-back waiting for data, we should finish our response only after all that data has arrived. Hence we also added a callback for the end event of the request object and in that callback we print to the console all the data that was sent by the client, and finish our response by calling its end method.

Let's try this

$ node examples/node/http_client_request_post.js 
Browse to http://127.0.0.1:8081

The regular GET request:

$ curl http://127.0.0.1:8081/

worked as earlier:

GET
{ 'user-agent': 'curl/7.37.1',
  host: '127.0.0.1:8081',
  accept: '*/*' }
/

the GET request with the path and the parameters:

$ curl http://127.0.0.1:8081/some/path?field=value

worked as well:

GET
{ 'user-agent': 'curl/7.37.1',
  host: '127.0.0.1:8081',
  accept: '*/*' }
/some/path?field=value

The question how does the POST request behave:

$ curl --data "field=value" http://127.0.0.1:8081/

and the result on the console is:

POST
{ 'user-agent': 'curl/7.37.1',
  host: '127.0.0.1:8081',
  accept: '*/*',
  'content-length': '11',
  'content-type': 'application/x-www-form-urlencoded' }
/
field=value

So indeed, this managed to collect the data that was sent by the client.

Comments

Hi Gabor, It is actually a beautiful tuto, I really appreciate. But i would say, it is seem like to be a node http server tuto rather then a node http client.

---

Its a great example to see how the legacy version of node consumed data. Thanks a lot