Introduction
The appeal of CoffeeScript
Prerequisites
Installation
The compiler
REPL
Simple web example
Conclusion
Downloadable resources
Related topics
Comments

Your first cup of CoffeeScript, Part 1

Getting started

Michael Galpin
Published on December 06, 2011

Content series:

This content is part # of # in the series: Your first cup of CoffeeScript, Part 1

https://www.ibm.com/developerworks/library/?series_title_by=**auto**

Stay tuned for additional content in this series.

This content is part of the series:Your first cup of CoffeeScript, Part 1

Stay tuned for additional content in this series.

The CoffeeScript programming language is built on top of JavaScript, and it compiles into efficient JavaScript that you can run in a web browser or use with technologies such as Node.js for server applications. The compilation is usually straightforward, and the JavaScript that is produced is consistent with many best practices. In this article, learn about the features of the CoffeeScript programming language. After you install CoffeeScript and run the compiler, you'll walk through a simple example using CoffeeScript in a web page.

Download the source code used in this article.

The appeal of CoffeeScript

Frequently used abbreviations

NPM: Node Package Manager
REPL: Read-Evaluate-Print-Loop

It's easy to argue that JavaScript is the most important programming language these days. It is the language of browsers, and is increasingly found in desktop and mobile applications. With the growing popularity of Node.js, JavaScript has become a viable option for server and system applications. Some developers vehemently resist JavaScript largely because of its inconsistent syntax and quirky implementations. However, the quirky implementations have subsided as more standardization has come to JavaScript virtual machines. The inconsistent syntax may be addressed somewhat by the next evolution of JavaScript: the ECMAScript.next, an emerging standard that has been heavily influenced by CoffeeScript. But, until a new standard is agreed upon and implemented by the popular virtual machines, there's a lot of room for improvement of JavaScript's syntax.

If you're writing for a JavaScript runtime, CoffeeScript is an enticing option. From a syntax perspective, JavaScript is quite the hodge-podge. It has many features of functional programming languages and is heavily influenced by Scheme, in particular. However, Scheme is a very simple syntax built on s-expressions. JavaScript shares many of the concepts in Scheme, but not its syntax. Instead, JavaScript has a C-like syntax. The result is a language with functional concepts, but a verbose syntax that has no natural constructs for expressing these concepts. For example, JavaScript allows for higher-order functions, such as functions whose input parameters include other functions. This is both useful and powerful, and is a feature missing from many languages. JavaScript can have not-so-elegant syntax, though, as shown in Listing 1.

Listing 1. Ugly JavaScript

pmb.requestPaymentInfo('type', function(info){
    $('result').innerHTML = info.name;
});

There is a lot of boilerplate in the example—parentheses, commas, curly braces, semicolons, and language keywords that are not really necessary.

JavaScript's primary use is as a client-side web application language. Desktop and mobile application frameworks such as Cocoa, Windows® Forms, and Android are all object-oriented. The object-oriented paradigm is not perfect for everything, but it's a nice fit for applications with a graphical user interface. JavaScript is also an object-oriented language complete with inheritance, but it is prototypal—not a class-based language like the languages used by most application frameworks. Thus, application programming in JavaScript can be very cumbersome.

CoffeeScript addresses the JavaScript pain points. CoffeeScript:

Provides a simpler syntax that reduces boilerplate, such as parentheses and commas
Uses whitespace as a way to organize blocks of code
Provides simple syntax for expressing functions
Supplies class-based inheritance (which is optional, but can be very useful when doing application development)

You might speculate that CoffeeScript, with an abstracted syntax, could have some disadvantages compared to JavaScript. For example, is CoffeeScript much slower than JavaScript, or does it require a heavy runtime library? In actuality, CoffeeScript compiles into clean, efficient JavaScript. You can always see exactly what it is compiling to, so you can be confident there is nothing excessive being introduced. And, because CoffeeScript compiles into fully functional JavaScript, there's no need for any type of runtime library. CoffeeScript gives you a syntax that lets you fully exploit the power of JavaScript, with minimal runtime overhead.

Prerequisites

As mentioned, you can use CoffeeScript to write server and system applications that run on top of Node.js. The relationship between CoffeeScript and Node.js runs deeper, though. To install CoffeeScript you need to first install Node.js because:

CoffeeScript is distributed as a Node.js package using node's package manager, NPM.
CoffeeScript must be compiled. Its compiler is actually written in CoffeeScript and, thus, requires a JavaScript runtime to do its compiling. The V8 JavaScript virtual machine at the heart of Node.js is perfect for the task.

To follow along with the example in this article, you need to install Node.js.

Installation

Have you ever wished you could run JavaScript from the command line? I haven't, but maybe CoffeeScript will change that. With Node.js, you can run JavaScript from the command line or as part of an executable script. This key feature of Node.js allows CoffeeScript code to be executed on the command line, providing the runtime needed for the CoffeeScript compiler (which is written in CoffeeScript).

The very first step is to install Node.js. You have several options for installation; you could compile the source code or run one of the installers that are available for various systems. Run node -v from the command line to confirm that Node.js is installed and on your path.

With Node.js you get an extra bonus: the node package manager (NPM). After you've run npm -v from the command line to confirm that NPM is installed and on your path, you can use NPM to install CoffeeScript as follows.

Run npm install --global coffee-script from the command line.
The --global flag makes CoffeeScript available system-wide and not just for a specific project.
The npm command should output something like /usr/bin/coffee -> /usr/lib/node_modules/coffee-script/bin/coffee.
NPM creates a shortcut in /usr/bin, so now the coffee executable is on the correct path. This is the CoffeeScript compiler and interpreter.
To verify that the coffee executable is on the path, run coffee -v from the command line.

There's one last step to ensure that the CoffeeScript environment is set up properly. To make CoffeeScript available to any Node.js processes that you start, you need to add it to what Node.js calls its NODE_PATH. Node.js will search NODE_PATH for modules (libraries) when it encounters unrecognized functions.

For the example in this article, you'll use Node.js mainly as a runtime for the CoffeeScript executables. The easiest approach is to simply add all of the NPM modules to NODE_PATH. To find where the NPM modules are located, enter npm ls -g. You need to add an environment variable pointing NODE_PATH to this location. For example, if npm ls -g prints /usr/lib, the modules are located in /usr/lib/node_modules. To set a NODE_PATH environment variable, run export NODE_PATH=/usr/lib/node_modules.

You could streamline things even more by putting the previous command in your startup script (~/.bash_profile, for example). To verify the changes, start a Node.js shell by executing Node and then type require('coffee-script'). The Node.js shell should load the CoffeeScript library. If it works, your CoffeeScript environment is good to go. You can now start exploring CoffeeScript by starting with the compiler.

The compiler

Running the CoffeeScript compiler is as easy as entering coffee -c, which launches the CoffeeScript read-evaluate-print-loop (REPL). To execute the compiler, you need to pass it a CoffeeScript file that you want to compile. Create a file called cup0.coffee and paste the contents of Listing 2 into the file.

Listing 2. Cup 0

for i in [0..5]
    console.log "Hello #{i}"

You can probably guess what the two lines of code in Listing 2 will do. Listing 3 shows the output of running coffee cup0.coffee.

Listing 3. Running your first CoffeeScript

$ coffee cup0.coffee 
Hello 0
Hello 1
Hello 2
Hello 3
Hello 4
Hello 5

To get a better idea of what is happening, try running the compiler. Enter coffee -c cup0.coffee, which creates a file named cup0.js. Listing 4 shows the contents of cup0.js.

Listing 4. Cup 0 JavaScript

(function() {
    var i;
    for (i = 0; i <= 5; i++) {
        console.log("Hello " + i);
    }
}).call(this);

A benefit of CoffeeScript is that, even though it provides a syntax that's more elegant than JavaScript, it compiles into very simple, logical JavaScript. You might wonder why all of the code is wrapped in a function. It's because JavaScript only supports function-level scoping. By wrapping everything in a function, you make sure that the variable is only scoped to that function and does not become global (or change an existing global variable).

Open a new file called cup1.coffee and enter the more complicated code in Listing 5.

Listing 5. Cup 1

stdin = process.openStdin()
stdin.setEncoding 'utf8'

stdin.on 'data', (input) -> 
    name = input.trim() 
    process.exit() if name == 'exit' 
    console.log "Hello #{name}"
    console.log "Enter another name or 'exit' to exit"
console.log 'Enter your name'

The program in Listing 5 prompts the user to enter their name and then greets them appropriately. JavaScript does not have any built-in libraries for reading from standard input, but Node.js does. This is another example of taking advantage of the CoffeeScript/Node.js symbiosis. In languages such as C, reading from standard input is a blocking call. No code can execute until the reading from standard input finishes. If you're familiar with Node.js, you know you cannot do things that way; Node.js does not allow for blocking I/O. Instead, you must register a callback with stdin.on.

Run coffee -c cup1.coffee to see, as in Listing 6, the JavaScript that the CoffeeScript compiler produces.

Listing 6. Cup 1 JavaScript

(function() {
    var stdin;
    stdin = process.openStdin();
    stdin.setEncoding('utf8');
    stdin.on('data', function(input) {
        var name;
        name = input.trim();
        if (name === 'exit') { 
            process.exit();
        }
        console.log("Hello " + name);
        return console.log("Enter another name or 'exit' to exit");
    });
   console.log('Enter your name');
}).call(this);

The stdin.on function uses a typical event binding format. You specify the kind of event ('data') you want to listen for, then give it a callback function to be executed when the event is fired. In the compiled JavaScript, you see typical verbose JavaScript for creating an inline function and passing it to another function. Compare this to the equivalent CoffeeScript. Does it make you miss all of those parentheses, curly braces, semicolons, and keywords?

Now that you know how to compile CoffeeScript programs, the next section looks at one of the most useful features for learning CoffeeScript: the REPL.

REPL

REPL is a standard tool found in many programming languages, particularly of the functional variety. REPL is the equivalent of Ruby's IRB. Simply entering coffee will launch the CoffeeScript REPL. For example, experiment with this CoffeeScript feature and solve simple problems, as shown in Listing 7.

Listing 7. Using the REPL

coffee> nums = [1..10]
[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 ]
coffee> isOdd = (n) -> n % 2 == 1
[Function]
coffee> odds = nums.filter isOdd
[ 1, 3, 5, 7, 9 ]
coffee> odds.reduce (a,b) -> a + b
25

Each time you enter an expression into the REPL, it will evaluate the expression, print the result, and then wait for the next expression. The example defines a variable called nums with a range from 1 to 10. The REPL prints out the value of the variable that was just defined. This feature can be immediately useful. Perhaps you don't remember if the range you defined is inclusive (includes the last number—in this case, 10)—or exclusive. The REPL shows you that 10 is included, so it is an inclusive range. If you wanted an exclusive range, just use nums = [1...10].

The isOdd function is defined next. CoffeeScript has very concise syntax for functional declarations, which is a nice feature given the functional nature of JavaScript. In the example, the REPL simply shows [Function] to let you know that the isOdd variable is equal to a function. A new variable, odds, is then declared. You get the value of odds by invoking the filter function on nums and passing in isOdd. This produces a new array whose elements must produce true when passed to isOdd. The reduce function is invoked on odds. Passing in a function adds each value of the array to the previous sum, which sums up the values in the array, and the REPL displays the sum of 25.

The next section covers a topic near and dear to any JavaScript developer: scripting in the browser.

Simple web example

You've seen how you can write CoffeeScript files and compile them into JavaScript, which, of course, can then be used in a web application. For development purposes, there is an easier way. The CoffeeScript compiler can run inside the browser, allowing you to use CoffeeScript directly in a web page. However, if you're building a high performance web application, this is not the advised way to use CoffeeScript. The CoffeeScript compiler is a large file; having to compile your CoffeeScript on the fly definitely slows things down. CoffeeScript does provide an easy way to develop your application with an obvious "path to production."

CoffeeScript is not a JavaScript toolkit or framework. It is a programming language, and, thus, does not include a lot of DOM-related convenience functions. However, you can use CoffeeScript with your favorite toolkit. The most common combination is to use it with jQuery. Listing 8 shows a simple web page that uses jQuery and CoffeeScript.

Listing 8. CoffeeScript in a web page

<html>
<head>
    <script src="//ajax.googleapis.com/ajax/libs/jquery/1.4.2/jquery.min.js">
</script>
    <script
 src="//jashkenas.github.com/coffee-script/extras/coffee-script.js">
 </script>
    <script type="text/coffeescript">
        gcd = (a,b) -> if (b==0) then a else gcd(b, a % b)
        $("#button").click -> 
            a = $("#a").val() 
            b = $("#b").val() 
            $("#c").html gcd(a,b)
    </script>
</head>
<body>
    A: <input type="text" id="a"/><br/>
    B: <input type="text" id="b"/><br/>
    <input type="button" value="Calculate GCD" id="button"/> <br/>
    C: <span id="c"/>
</body>
</html>

The web page uses the Google Ajax library to load jQuery. It loads the CoffeeScript compiler library from CoffeeScript creator Jeremy Ashkenas' Github repository (see Related topics). The code then includes a script block. Instead of it being of type text/javascript, the script block is of type text/coffeescript, which is how the CoffeeScript compiler knows to compile the script contents. The script then creates a function called gcd that calculates the greatest common divisor of two integers. jQuery is used to create a click handler for the button on the page. In this handler you get the values of the two text inputs and pass those to the gcd function. The result is written back to the web page. Functions such as $(), val(), and html() are jQuery functions, but they can be easily used with CoffeeScript and take advantage of CoffeeScript's clean syntax.

Conclusion

In this article, you got a quick tour of CoffeeScript. With your development environment up and running, you can now explore CoffeeScript using the REPL. You learned to use the compiler to see what kind of JavaScript it produces, and also learned how to write CoffeeScript and run it directly in your web pages. The examples provided a taste of CoffeeScript's syntax, though often without a lot of explanation.

Part 2 in this series will dive deeper into the details of CoffeeScript key concepts.

Downloadable resources

PDF of this content
Article source code (cs1.zip | 1KB)

Contents