Parse HTML without xpath

I'm trying to create a simple tool to parse html files.

Specifically, I开发者_开发知识库 need it to get all the name attributes out of all the div tags.

My HTML string varies and I don't have any control over it, so if I try and use xpath I tend to get errors as the HTML is not 100% written correctly.

Any ideas?

Thanks,

There is also a great class called PHP Simple HTML DOM Parser on http://simplehtmldom.sourceforge.net/

Works fine with invalid HTML, but needs a lot of memory for parsing long html-files.