Need a spider library

[Laszlo Zsolt Nagy]

  Hi All,

I'm writting a spider program. I need to go to serveral URLs and extract 
information from the HTML source. Including links.
I was using FancyURLOpener and my own function that extracts the links 
from a HTML page. The problem is that I always
need to change it. This is because some sites use lower case tag names, 
others upper case tag names. Some of them use
href="page.html" others do it without the quotation   href=page.html but 
I could even find unclosed quotations <a href="page.html>
double opened and unclosed <a tags etc. There are many kinds of 
malformed HTML pages out there and it seems I'm not capable
of handling all of them. The question: is there a good library for 
Python for extraction links and images out  of (possibly malformed)
HTML soucre code? Like the "references" function in Lynx. I need to 
handle relative and absolute references and I need to know the
anchor text too and the position of the anchor inside the HTML source file.

For example this malformed link:

<a href="page.html>Sample link</a>

could be converted to:

['page.html','http://samplesite.current_location/page.html','Samle link']

Thanks in advance

   Les