fix issue#63: make dependency markit upgrade to 0.2.0 and fixed post title parsing problem

[qhhonx]

markit v0.2.0 does generate head tag code snippet with id attribute which leads to
html.match(/<h1>(.*?)<\/h1>/); can not match anything.

[lepture]

@xohozu Can you fix the test cases? It seems that shouldjs has no include method now.

[qhhonx]

@lepture I checked the newest api document of shouldjs and replace the obsolete include method. However, the tests also can not pass because when nico used [email protected] and it generated some different code snippets compared with using [email protected].

[email protected]

<h1 id="h1-1">h1 1</h1><p>hello world</p>
<div class="nico-insert-code"><div>html</div></div><div class="highlight"><pre><code class="xml"><span class="tag">&lt;<span class="title">div</span>&gt;</span>html<span class="tag">&lt;/<span class="title">div</span>&gt;</span></code></pre></div><script>
seajs.use("foo")
</script>

<h1 id="h1-2">h1 2</h1><h2 id="h2">h2</h2><div class="nico-iframe">
<iframe src="iframe--1.html" allowtransparency="true" 
frameborder="0" scrolling="0" height="100"></iframe></div><div class="highlight"><pre><code class="xml">iframe content</code></pre></div>

[email protected]

<h1 id="h1-1">h1 1</h1>
<p>hello world</p>
<pre><code class="lang-html+">&lt;div&gt;html&lt;/div&gt;
</code></pre>
<script>
seajs.use("foo")
</script>

<h1 id="h1-2">h1 2</h1>
<h2 id="h2">h2</h2>
<pre><code class="lang-iframe:100+">iframe content
</code></pre>

I can not solve these problems in a short time because this is the first time I read nodejs codes.
Hope you can fix it.