learnyounode Lesson 11 – HTTP File Server

For this lesson we need to create an HTTP server that serves the same text file each time the server receives a request.  Just like in the previous lesson, the server should listen on the port that is provided by the first argument to the program (process.argv[2]).  A location to the file will be provided as the second argument to the program.  Unlike the previous lesson, we need to create an HTTP server rather than a TCP server.

This lesson requires that we use the fs.createReadStream() method to stream the file contents to our server’s response rather than using the fs.readFile() method.  What’s the difference?  The fs.readFile() method will read the entire file into memory before sending it to the response, while the fs.createReadStream() method will stream the file contents to the response as it is read.  This may be faster in some cases, and use less memory.

By creating a read stream, we can use the pipe() method to stream our files contents from the source (the file) to the destination (the response).  The signature for the pipe method is source.pipe(destination), although you will see it represented as src.pip(dst) in the tutorial hint.  The bottom line is that we can connect a file system stream to a response stream.

Solution

The solution I came up with is very similar to the official solution.  I am passing the http and fs modules to the variables http and fs.  In addition, I am passing the first and second arguments (process.argv[2] and process.argv[3]) to my program to the variables port and file.  It is not necessary to pass either arguments to a variable.  However, I find that it is helpful for bringing context and readability to my program by naming the purposes of those arguments.

After declaring my variables, I am using the http.createServer() method and passing a callback function as an argument.  The callback function takes two arguments called request and response.  Inside of my callback function, I am using the fs.createReadStream() method to to stream my file to memory.  Then I am passing that stream to the response object of my callback function with the pipe() method.  Finally, I am listening on the port that is passed to my program via process.argv[3] with the http.listen() method.  Rather than calling method below createServer(), I am appending it to the end of the http.createServer method with the signature http.createServer().listen().  Inside of the listen() method, I am also passing a callback function as an argument.  When a connection is made, my callback function logs the url of my server to the console and tells the user that the server is listening.

It is worth mentioning that I often times see folks placing the ‘server is listening’ message outside of a callback function.  As far as I know, this is not the correct way to go about it because the server is listening message may be logged to the console before the server is actually listening.  It is best to place this message in a callback so that your end user knows that the server is actually listening.

Official Solution

Click to show/hide solution



The official solution differs only slightly.  The key difference is that after the createServer() method is called, the writeHead() method is called inside of a callback function that is passed to the createServer() method as an argument.  The data that is written to the head is the http success code (200), as well as a JSON object which contains the content type.  This really is the best practice.  If you wanted to you could take this a step further and also write the content length the the response head by using the length() method to calculate the length of your data.

Lesson 12

Contents