Downloading images with node.js

I'm trying to write a script to download images using node.js. This is what I have so far:

var maxLength = 10 // 10mb
var download = function(uri, callback) {
  http.request(uri)
    .on('response', function(res) {
      if (res.headers['content-length'] > maxLength*1024*1024) {
        callback(new Error('Image too large.'))
      } else if (!~[200, 304].indexOf(res.statusCode)) {
        callback(new Error('Received an invalid status code.'))
      } else if (!res.headers['content-type'].match(/image/)) {
        callback(new Error('Not an image.'))
      } else {
        var body = ''
        res.setEncoding('binary')
        res
          .on('error', function(err) {
            callback(err)
          })
          .on('data', function(chunk) {
            body += chunk
          })
          .on('end', function() {
            // What about Windows?!
            var path = '/tmp/' + Math.random().toString().split('.').pop()
            fs.writeFile(path, body, 'binary', function(err) {
              callback(err, path)
            })
          })
      }
    })
    .on('error', function(err) {
      callback(err)
    })
    .end();
}

I, however, want to make this more robust:

  1. Are there libraries that do this and do this better?
  2. Is there a chance that response headers lie (about length, about content type)?
  3. Are there any other status codes I should care about? Should I bother with redirects?
  4. I think I read somewhere that binary encoding is going to be deprecated. What do I do then?
  5. How can I get this to work on windows?
  6. Any other ways you can make this script better?

Why: for a feature similar to imgur where users can give me a URL, I download that image, and rehost the image in multiple sizes.

I'd suggest using the request module. Downloading a file is as simple as the following code:

var fs = require('fs'),
    request = require('request');

var download = function(uri, filename, callback){
  request.head(uri, function(err, res, body){
    console.log('content-type:', res.headers['content-type']);
    console.log('content-length:', res.headers['content-length']);

    request(uri).pipe(fs.createWriteStream(filename)).on('close', callback);
  });
};

download('https://www.google.com/images/srpr/logo3w.png', 'google.png', function(){
  console.log('done');
});

Building on top of Cezary's solution: Probably you will need a callback too.

var fs = require('fs'),
    request = require('request');

var download = function(uri, filename, callback){

  request.head(uri, function(err, res, body){

    console.log('content-type:', res.headers['content-type']);
    console.log('content-length:', res.headers['content-length']);

    var r = request(uri).pipe(fs.createWriteStream(filename));
    r.on('close', callback);
  });
};

download('https://www.google.com/images/srpr/logo3w.png', 'google.png', function(){
    console.log('Done downloading..');
  });

Building on the above, if anyone needs to handle errors in the write/read streams, I used this version. Note the stream.read() in case of a write error, it's required so we can finish reading and trigger close on the read stream.

var download = function(uri, filename, callback){
  request.head(uri, function(err, res, body){
    if (err) callback(err, filename);
    else {
        var stream = request(uri);
        stream.pipe(
            fs.createWriteStream(filename)
                .on('error', function(err){
                    callback(error, filename);
                    stream.read();
                })
            )
        .on('close', function() {
            callback(null, filename);
        });
    }
  });
};

I ran into this problem some days ago, for a pure NodeJS answer I would suggest using Stream to merge the chunks together.

var http = require('http'),                                                
    Stream = require('stream').Transform,                                  
    fs = require('fs');                                                    

var url = 'http://www.google.com/images/srpr/logo11w.png';                    

http.request(url, function(response) {                                        
  var data = new Stream();                                                    

  response.on('data', function(chunk) {                                       
    data.push(chunk);                                                         
  });                                                                         

  response.on('end', function() {                                             
    fs.writeFileSync('image.png', data.read());                               
  });                                                                         
}).end();

The newest Node versions won't work well with binary strings, so merging chunks with strings is not a good idea when working with binary data.

*Just be careful when using 'data.read()', it will empty the stream for the next 'read()' operation. If you want to use it more than once, store it somewhere.