JavaScript compressor/beautifier on top of PARSE-JS

cl-uglify-js

<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
               "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"
lang="en" xml:lang="en">
<head>
<title>cl-uglify-js -- JavaScript compressor/beautifier for Common Lisp</title>
<meta http-equiv="Content-Type" content="text/html;charset=utf-8"/>
<meta name="generator" content="Org-mode"/>
<meta name="generated" content="2011-01-24 00:09:50 EET"/>
<meta name="author" content="Mihai Bazon"/>
<meta name="description" content="JavaScript compressor/beautifier for Common Lisp"/>
<meta name="keywords" content="javascript, js, parser, compiler, compressor, mangle, minify, minifier, common lisp"/>
<style type="text/css">
 <![CDATA[/*>
</style>
<link rel="stylesheet" type="text/css" href="docstyle.css" />
<script type="text/javascript">
<![CDATA[/*>
</script>

</head>
<body>
<div id="content">

<h1 class="title">cl-uglify-js &ndash; JavaScript compressor/beautifier for Common Lisp</h1>

<p>This is a Common Lisp version of <a href="http://github.com/mishoo/UglifyJS">UglifyJS</a>.  It works on data produced by
<a href="http://marijn.haverbeke.nl/parse-js/">parse-js</a> to generate a “minified” version of the code.  Currently
it can:
</p>
<ul>
<li>
reduce variable names (usually to single letters)
</li>
<li>
join consecutive <b>var</b> statements
</li>
<li>
resolve simple binary expressions
</li>
<li>
group most consecutive statements using the “sequence” operator (comma)
</li>
<li>
remove unnecessary blocks
</li>
<li>
convert IF expressions in various ways that result in smaller code
</li>
<li>
remove some unreachable code

</li>
</ul>

<p>It's faster than YUI Compressor, Google Closure <i>and</i> UglifyJS, and almost
as good as (and a bit safer than) Google Closure in terms of compressed code
size.
</p>
<p>
See the <a href="http://github.com/mishoo/UglifyJS">UglifyJS</a> page for more information.
</p>

<div id="table-of-contents">
<h2>Table of Contents</h2>
<div id="text-table-of-contents">
<ul>
<li><a href="#sec-1">1 Dependencies </a></li>
<li><a href="#sec-2">2 API </a>
<ul>
<li><a href="#sec-2_1">2.1 ast-squeeze (ast &amp;key (sequences T) (dead-code T)) </a></li>
<li><a href="#sec-2_2">2.2 ast-mangle (ast &amp;key toplevel) </a></li>
<li><a href="#sec-2_3">2.3 ast-gen-code (ast &amp;key (beautify T) (indent-level 4) (indent-start 0) quote-keys) </a></li>
<li><a href="#sec-2_4">2.4 split-code (code &amp;optional (maxlen (* 32 1024))) </a></li>
</ul>
</li>
<li><a href="#sec-3">3 Compress one file </a></li>
<li><a href="#sec-4">4 License </a></li>
</ul>
</div>
</div>

<div id="outline-container-1" class="outline-3">
<h3 id="sec-1"><span class="section-number-3">1</span> Dependencies </h3>
<div class="outline-text-3" id="text-1">


<ul>
<li>
<a href="http://marijn.haverbeke.nl/parse-js/">parse-js</a>
</li>
<li>
<a href="http://weitz.de/cl-ppcre/">cl-ppcre</a>
</li>
<li>
<a href="http://www.cliki.net/PARSE-NUMBER">parse-number</a>
</li>
<li>
<a href="http://common-lisp.net/project/iterate/">iterate</a>

</li>
</ul>

<p>I only tested it on SBCL.
</p>
</div>

</div>

<div id="outline-container-2" class="outline-3">
<h3 id="sec-2"><span class="section-number-3">2</span> API </h3>
<div class="outline-text-3" id="text-2">


<p>
The following functions are exported:
</p>

</div>

<div id="outline-container-2_1" class="outline-4">
<h4 id="sec-2_1"><span class="section-number-4">2.1</span> ast-squeeze (ast &amp;key (sequences T) (dead-code T)) </h4>
<div class="outline-text-4" id="text-2_1">


<p>
Applies various compression techniques.  It expects an AST (as returned by
<a href="http://marijn.haverbeke.nl/parse-js/">parse-js</a>) and returns a new, compatible AST (possibly sharing structure with
the original one!).
</p>
<p>
Optional keyword arguments:
</p>
<ul>
<li>
<code>sequences</code> (default <code>T</code>) &mdash; set this to <code>NIL</code> to disable grouping
consecutive statements into a sequence using the “comma operator”.
</li>
<li>
<code>dead-code</code> (default <code>T</code>) &mdash; if you pass <code>NIL</code> it will not attempt to
remove unreachable code.

</li>
</ul>

<p>When it encounters unreachable code and <code>dead-code</code> is <b>true</b>, this function
will <code>warn</code> about it.  The warnings go to <code>*error-output*</code>, so rebind that
if you want to catch the messages.
</p>
</div>

</div>

<div id="outline-container-2_2" class="outline-4">
<h4 id="sec-2_2"><span class="section-number-4">2.2</span> ast-mangle (ast &amp;key toplevel) </h4>
<div class="outline-text-4" id="text-2_2">


<p>
“Mangles” variable names (renames all variables to shorter version).  This
function is careful not to affect the semantics of the code.  It will avoid
renaming undeclared variables (which could possibly be defined in some other
script), and avoid renaming names that are under the influence of a <b>with</b>
block, or within the context of an <b>eval</b> call.
</p>
<p>
Optional keyword arguments:
</p>
<ul>
<li>
<code>toplevel</code> (default <code>NIL</code>) &mdash; pass <b>true</b> here if you want to mangle the
toplevel scope.  By default we don't.

</li>
</ul>

<p>Note that this function returns a somewhat incompatible AST.  Literal names
(which are normally strings) are replaced with lambda-s that would return
the mangled version.  <code>ast-gen-code</code> knows how to deal with this.
</p>
</div>

</div>

<div id="outline-container-2_3" class="outline-4">
<h4 id="sec-2_3"><span class="section-number-4">2.3</span> ast-gen-code (ast &amp;key (beautify T) (indent-level 4) (indent-start 0) quote-keys) </h4>
<div class="outline-text-4" id="text-2_3">


<p>
Given an abstract syntax tree, this function returns the corresponding
JavaScript code (as a string).
</p>
<p>
Optional keyword arguments:
</p>
<ul>
<li>
<code>beautify</code> (default <b>true</b>).  By default ast-gen-code returns nicely
indented code.  If you want to compress, pass <code>:beautify NIL</code>.

</li>
</ul>

<p>The other arguments only make sense if <code>beautify</code> is <b>true</b>:
</p>
<ul>
<li>
<code>indent-level</code>: number of spaces to use for indentation.  Note that
case/default statements are indented to half of this number, so better
pass an even one.

</li>
<li>
<code>indent-start</code>: the whole code will be indented by this number of spaces
(default 0).

</li>
<li>
<code>quote-keys</code>: by default, we only quote keys that cannot be used otherwise
(i.e. reserved words such as "while").  Pass this <b>true</b> to quote all keys
regardless.
</li>
</ul>


</div>

</div>

<div id="outline-container-2_4" class="outline-4">
<h4 id="sec-2_4"><span class="section-number-4">2.4</span> split-code (code &amp;optional (maxlen (* 32 1024))) </h4>
<div class="outline-text-4" id="text-2_4">


<p>
Given `code' (a string) it will split it by adding a newline every `maxlen'
(or so) characters.  I found both Firefox and Chrome croak with weird errors
when the whole code was on a single 680K line, so even if it adds a few more
bytes this step should be done for safety.  The default `maxlen' is 32K.
</p>
</div>
</div>

</div>

<div id="outline-container-3" class="outline-3">
<h3 id="sec-3"><span class="section-number-3">3</span> Compress one file </h3>
<div class="outline-text-3" id="text-3">


<p>
To compress one file, you would do something like this:
</p>



<pre class="src src-lisp">(ast-gen-code
  (ast-mangle
    (ast-squeeze
      (<span style="color: #afeeee; font-weight: bold;">with-open-file</span> (in <span style="color: #87cefa;">"/path/to/file.js"</span>)
        (parse-js:parse-js in)))) <span style="color: #7fffd4;">:beautify</span> nil)
</pre>



<p>
Some simplified API will be available at some point, though I'm not sure it
would be really useful.
</p>
</div>

</div>

<div id="outline-container-4" class="outline-3">
<h3 id="sec-4"><span class="section-number-3">4</span> License </h3>
<div class="outline-text-3" id="text-4">





<pre class="example">Copyright 2010 (c) Mihai Bazon &lt;mihai.bazon@gmail.com&gt;

This software is provided 'as-is', without any express or implied
warranty. In no event will the authors be held liable for any
damages arising from the use of this software.

Permission is granted to anyone to use this software for any
purpose, including commercial applications, and to alter it and
redistribute it freely, subject to the following restrictions:

1. The origin of this software must not be misrepresented; you must
   not claim that you wrote the original software. If you use this
   software in a product, an acknowledgment in the product
   documentation would be appreciated but is not required.

2. Altered source versions must be plainly marked as such, and must
   not be misrepresented as being the original software.

3. This notice may not be removed or altered from any source
   distribution.
</pre>


</div>
</div>
<div id="postamble">
<p class="author"> Author: Mihai Bazon
</p>
<p class="date"> Date: 2011-01-24 00:09:50 EET</p>
<p class="creator">HTML generated by org-mode 7.01trans in emacs 23</p>
</div>
</div>
</body>
</html>

cl-uglify-js

Upstream URL

Author

License

cl-uglify-js

Authors:

Dependencies:

Dependencies (4)

Dependents (1)