Modern Python Standard Library Cookbook

There's more...

The proposed recipe shows how to use HTMLParser to adapt the XML parsing utilities to cope with HTML, which is more flexible in rules when compared with XML.

While this solution handles mostly commonly written HTML, it won't cover all possible cases. HTML supports some oddities that are sometimes used, such as attributes without any value:

<input disabled>

Or attributes without quotes:

<input type=text>

And even some attributes with content but without any closing tag:

<li>Item 1
<li>Item 2

Even though most of these formats are supported, they are rarely used (with maybe the exception of attributes without any value, which our parser will just report as having a value of None), so in most cases, they won't cause trouble. But ...