CODEWITHSUNDEEP
stringnode.jsbinarybuffer
CuriousMind
CuriousMind

Reputation: 34145

Why do we need buffers in node.js?

To display the contents of the file in browser using node, I first tried this:

var express = require('express')
  , fs  = require('fs')
  , app = express()
  , port = process.env.PORT || 5000;

app.use(express.logger());

var data = fs.readFileSync('index.html');


app.get('/', function(req, res){
  res.send(data);
});

app.listen(port, function(){
  console.log('Listenting on ' + port);
});

However, this didn't work as expected. Browser actually offered to contents of index.html as file that I first have to download & then manually open that in text editor. then, I tried using buffer to do the same thing:

var express = require('express')
  , fs  = require('fs')
  , app = express()
  , port = process.env.PORT || 5000;

app.use(express.logger());

var data = new Buffer(fs.readFileSync('index.html'));


app.get('/', function(req, res){
  res.send(data.toString('utf-8'));
});

app.listen(port, function(){
  console.log('Listenting on ' + port);
});

This works as expected & contents of index.html were display inside the browser window. However, In My limited userstanding about Node.js. Buffer class is a global type for dealing with binary data. However the index.html file didn't contain any binary data. so my questions are:

  1. In above code, why did the contents of index.html were offered as file download(without buffer) vs being displayed directly in browser(with buffer)?
  2. Why do I need to use buffer when the file contains only string data, no binary?
  3. What are the offer places where buffers are used?
  4. If I want to learn more about buffers & there use in Node. where do I go?

Upvotes: 4

Views: 2779

Answers (1)

AndyD
AndyD

Reputation: 5385

The correct way to return a file would be to use an async method instead of a sync method.:

 app.get('/', function(req, res){
   res.writeHead(200, {
    'Content-Type': 'text/plain' // set to whatever mime type you need. 
  });
  fileSystem.createReadStream(filePath).pipe(res);
});

If you're going to serve static files with express then you should use the static middleware.

To find out the differences in your example, have a look at the content-type being returned. That's what would determine what the browser does with your response.

In the first case, Express.js would have seen a Buffer and decided to return the data as content-type "application/octet-stream". Your browser would not know what to do with this and return it as a download file.

In the second case, Express.js would have seen a String and decided to return the data as content-type "text/plain". Your browser knows how to deal with that and would show it to you.

I don't think your question has much to do with buffers in node.js but more to do with how Express.js res.send handles various input types.

If you're new to Node then you don't need to know about Buffers right away I think. Otherwise have a look at the docs: http://nodejs.org/api/buffer.html#buffer_buffer

Upvotes: 3

Related Questions