node/node.html


<html>
<style>
body {
  background: #22252a;
  color: #eee;
  font-size: 16pt;
  line-height: 150%;
  font-family: times, Times New Roman, times-roman, georgia, serif; 
}
#content {
  max-width: 30em;
  margin: 0 0 5em 10em;
}
#toc {
  position: fixed;
  top: 2em;
  left: 0;
  width: 10em;
}
#toc ol {
  list-style: none;
  margin: 0;
  padding: 0;
  padding-left: 1em;
}
#toc ol li {
  margin: 0;
  padding: 0;
}
#toc a { color: #777; }

h1, h2, h3, h4 { color: #bbb; }

h1 {
  margin: 2em 0;
  padding: 0px 0px 0px 0px;
  font-size: 51px;
  line-height: 44px;
  font-weight: bold;
}
h1 a { color: inherit; }

h2 {
  font-size: 30px;
  line-height: inherit;
  font-weight: bold;
  margin: 2em 0;
}

h3 {
  margin: 2em 0;
  font-size: 20px;
  line-height: inherit;
  font-weight: bold;
}

h4 {
  margin: 1em 0;
  font-size: inherit;
  line-height: inherit;
  font-weight: bold;
}

pre, code {
  font-family: monospace;
  font-size: 13pt;
  color: #aaf;
}

pre {
  padding-left: 2em;
}

dl {
}

dt {
  color: #aaf;
  font-weight: bold;
}

dd {
  margin: 1em 0;
  margin-left: 1em;
}

a { color: #cd5; text-decoration: none; }
a:hover { text-decoration: underline; }

.highlight {
  background: #733;
  padding: 0.2em 0;
}

</style>
<title>node.js</title>
<body>
<div id="toc">
  <ol>
    <li><a href="#motivation">Motivation</a></li>
    <li><a href="#benchmarks">Benchmarks</a></li>
    <li><a href="#download">Download</a></li>
    <li><a href="#install">Build</a></li>
    <li><a href="#api">API</a>
      <ol>
        <li><a href="#timers">Timers</a>
        <li><a href="#files">File System</a>
        <li><a href="#tcp">TCP</a>
        <li><a href="#http">HTTP</a>
          <ol>
          <li><a href="#http_server">Server</a>
          <li><a href="#http_client">Client</a>
          </ol>
        <li><a href="#modules">Modules</a>
      </ol>
    </li>
  </ol>
</div>
<div id="content">

<h1><a href="http://tinyclouds.org/node">Node</a></h1>

<p id="introduction"> Node is a purely asynchronous I/O framework for <a
  href="http://code.google.com/p/v8/">V8 javascript</a>. 
  
<p>This is an example of a web server written with Node which responds with
"Hello World" after waiting two seconds:
<pre class="sh_javascript">
new node.http.Server(function (msg) {
  setTimeout(function () {
    msg.sendHeader(200, [["Content-Type", "text/plain"]]);
    msg.sendBody("Hello World");
    msg.finish();
  }, 2000);
}).listen(8000, "localhost");
</pre>

<p> This script can handle hundreds of concurrent requests while using
little CPU or memory&mdash;<a href="#benchmarks">see benchmarks</a>.
Check out <a href="#api">the documentation</a> for more examples.

<p> Node is free to <a href="#download">download</a>, <a
  href="#api">use</a>, and <a href="#modules">build upon</a>.</p>

<h2 id="motivation">Motivation</h2>

<h3>Evented Programming Makes More Sense</h3>

difference between blocking/non-blocking design

<p> There are many methods to write internet servers but they can
fundamentally be divided into two camps: evented and threaded; non-blocking
and blocking. A blocking server accepts a connection and launches a new
thread to handle the connection. Because the concurrency is handled by
the thread scheduler, a blocking server can make function calls which
preform full network requests.

<pre class="sh_javascript">var response = db.execute("SELECT * FROM table");
// do something</pre> 

<p> An evented server manages its concurrency itself. All connections
are handled in a single thread and callbacks are executed on certain
events: "socket 23 is has data to read", "socket 65's write buffer is
empty". An evented server executes small bits of code but never
<i>blocks</i> the process. In the evented world callbacks are used
instead of functions

<pre class="sh_javascript">db.execute("SELECT * FROM table", function (response) {
  // do something
});</pre>


<p><a href="http://duartes.org/gustavo/blog/post/what-your-computer-does-while-you-wait">I/O latency</a>
<pre>
l1 cache ~ 3
l2 cache ~ 14
     ram ~ 250
    disk ~ 41000000
 network ~ 240000000
</pre>

<p>purely evented interfaces rule out a lot of stupidity

<h3>Evented programs are more efficient</h3>
  <ol>
  <li>pthread stack size
   2mb default stack size on linux (1mb on windows, 64kb on FreeBSD)
   of course this is adjustable 
  <li>context switching benchmark
  <li>Apache vs. Nginx 
  <li>event machine vs mongrel (neverblock)
  </ol>
<h3>The appropriateness of Javascript</h3>
  <ol>
  <li>No I/O 
    <p> Javascript is without I/O. In the browser the DOM provides I/O,
    but non-browser javascript interpreters have only non-standardized
    functions to allow them print to console or access the network.
  <li>No Threads
  <li>Good compiler
  <li>Universality of the language
    <p> Contemporary computer infrastructure has two irreplaceable
    languages: C and Javascript. C is the language of operating systems.
    POSIX, the universal operating system API, is defined in C. So while you
    can interface with operating systems in Java and Haskell, those
    languages access must make system calls in C.  Similarly, Javascript is
    the language of the web operating system.  In place of POSIX is the
    DOM. You can wrap Javascript, you can compile to Javascript, but in the
    end browsers must be interfaced with in Javascript. Portable low-level
    systems tend to be written in C and portable web-level systems are
    written in Javascript.
  </ol>

<h2 id="benchmarks">Benchmarks</h2>

<p> TODO

<h2 id="download">Download</h2>

<p> TODO

<h2 id="build">Build</h2>

<pre>configure
make
make install</pre>


<h2 id="api">Application Programming Interface</h2>

<p> The node executable should be given an argument pointing to a
javascript file. 

<h3 id="timers">Timers</h3>

<p>The timer API is the same as in the browser. The functions
<code>setTimeout, setInterval, clearTimeout, and clearInterval</code>

<h3 id="files">File System</h3>

<h3 id="tcp">TCP</h3>

<h3 id="http">HTTP (<code>node.http</code>)</h3>

<p> Node provides a fast web server and client interface. The interface is
rather low-level. Node, for example, will not parse
<code>application/x-www-form-urlencoded</code> message bodies&mdash;that is
for higher level code to manage. The interface does abstract the
Transfer-Encoding (i.e. chuncked or identity), Message boundarys, and
Keep-Alive connections. 

<h4 id="http_server">HTTP Server (<code>node.http.Server</code>)</h4>
<dl>
  <dt><code>new Server(request_handler, options)</code></dt>
  <dd> 
      <p>Creates a new web server. The <code>options</code> argument accepts
      the same values as the options argument for
      <code>node.tcp.Server</code> does. The options argument is optional. 

      <p>The <code>request_handler</code> is a function which is called on 
      each request with a <code>Message</code> object argument.
  </dd>

  <dt><code>server.listen(port, hostname)</code>
  <dd> 
    <p>Begin accepting connections on the specified port and hostname. If the
    hostname is omitted, the server will accept connections directed to any
    address.  
  </dd>

  <dt><code>server.close()</code>
  <dd>
    <p>Stops the server. Requests currently in progress will not be
    interrupted.
  </dd>
</dl>
    

<h4>HTTP Request Message (<code>node.http.Message</code>)</h4>

<p> This object is only created internally&mdash;not by the user. It is passed
as an argument to the <code>request_handler</code> function in a web server. 
<dl>
  <dt><code>msg.method</code>
  <dd>The request method as a string. Read only. Example: <code>"GET"</code>,
  <code>"DELETE"</code>.</dd>

  <dt><code>msg.uri</code>
  <dd>The request URI as a string. Read only.
  Example: <code>"/index.html?hello=world"</code>.</dd>

  <dt><code>msg.headers</code>
  <dd>The request headers expressed as an array of 2-element arrays. Read only.  
  Example:
<pre>
[ ["Content-Length", "123"]
, ["Content-Type", "text/plain"]
, ["Connection", "keep-alive"]
, ["Accept", "*/*"]
]
</pre>

  <dt><code>msg.http_version</code></dt>
  <dd>The HTTP protocol version as a string. Read only. Examples: <code>"1.1"</code>,
  <code>"1.0"</code>

  <dt><code>msg.connection</code></dt>
  <dd> A reference to the <code>node.tcp.Connection</code> object. Read
  only. Note that multiple messages can be sent on a single connection.
  </dd>

  <dt><code>msg.onBody</code></dt>
  <dd>Callback. Should be set by the user to be informed of when a piece
  of the message body is received. Example:
<pre>
msg.onBody = function (chunk) {
  puts("part of the body: " + chunk);
}
</pre>
  A chunk of the body is given as the single argument. The transfer-encoding
  has been removed. 
  <p>The body chunk is either a String in the case of utf8 encoding or an
  array of numbers in the case of raw encoding.

  <dt><code>msg.onBodyComplete</code></dt>
  <dd>Callback. Made exactly once for each message. No arguments. After
  <code>onBodyComplete</code> is executed <code>onBody</code> will no longer be called.
  </dd>

  <dt><code>msg.setBodyEncoding(encoding)</code></dt>
  <dd>
    Set the encoding for the request body. Either <code>"utf8"</code> or
    <code>"raw"</code>. Defaults to raw.
    <big>TODO</big>

  <dt><code>msg.sendHeader(status_code, headers)</code></dt>
  <dd>
    Sends a response header to the request. The status code is a 3-digit
    HTTP status code, like <code>404</code>. The second argument,
    <code>headers</code>, should be an array of 2-element arrays,
    representing the response headers. 

    <p>Example:
<pre>
var body = "hello world";
msg.sendHeader( 200
              , [ ["Content-Length", body.length] 
                , ["Content-Type", "text/plain"] 
                ]
              );
</pre>
    This method must only be called once on a message and it must be called
    before <code>msg.finish()</code> is called.
  </dd>

  <dt><code>msg.sendBody(chunk)</code></dt>
  <dd>
  This method must be called after <code>sendHeader</code> was called. It
  sends a chunk of the response body. This method may be called multiple
  times to provide successive parts of the body.
  </dd>

  <dt><code>msg.finish()</code></dt>
  <dd>
  This method signals that all of the response headers and body has been
  sent; that server should consider this message complete. 
  The method, <code>msg.finish()</code>, MUST be called on each response.

</dl>

<h3 id="modules">Modules</h3>

<p>Node has simple module loading.  Here is an example. This is the file
<code>foo.js</code>:
<pre>
include("mjsunit");

function onLoad () {
  assertEquals(1, 2);
}
</pre>
<p>Here the module <code>mjsunit</code> has provided the function
<code>assertEquals()</code>.

<p> The module file, <code>mjsunit.js</code>, must be in the same directory
as <code>foo.js</code> for <code>include()</code> to work. The
<code>include()</code> function will insert all the exported objects from the
module into the global namespace.

<p> Because file loading does not happen instantaneously, and because Node
has a policy of never blocking, the callback <code>onLoad()</code> is
provided to notify the user when all the exported functions are completely
loaded.

<p> To export an object, add to the special object <code class="highlight">exports</code>.
Let's look at how <code>mjsunit.js</code> does this

<pre>
function fail (expected, found, name_opt) {
  // ...
}

function deepEquals (a, b) {
  // ...
}

<span class="highlight">exports</span>.assertEquals = function (expected, found, name_opt) {
  if (!deepEquals(found, expected)) {
    fail(expected, found, name_opt);
  }
};
</pre>
<p> The functions <code>fail</code> and <code>deepEquals</code> are not
exported and remain private to the module.

<p> In addition to <code>include()</code> a module can use
<code>require()</code>. Instead of loading the exported objects into the
global namespace, it will return a namespace object. Again, the members of
the namespace object can only be guaranteed to exist after the
<code>onLoad()</code> callback is made. For example:
<pre>
var mjsunit = require("mjsunit");

function onLoad () {
  mjsunit.assertEquals(1, 2);
}
</pre>


</body>
</html>