Mechanize users - Jan 2008 - Quick and Easy Question

All,

Been a while since I''ve "mechanized" and I''m running
into an idiot simple
problem.  I''m looking for more detailed examples to refresh my memory.

My question is how to get a list of links off of a page based on information
in the img src attribute, specifically if the img tag contains a
"mp3.gif/jpg".
I, for the life of me, can''t seem to  remember how to do this.

And, as I go through this exercise, I am sure I will run into other issues
as well.  So, if anyone has any good sources for examples, I''d love to
see
them.

Thanks,

William
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://rubyforge.org/pipermail/mechanize-users/attachments/20080114/53f77921/attachment.html

Hey William!

On Mon, Jan 14, 2008 at 10:25:16AM -0500, William Flanagan
wrote:
> All,
> 
> Been a while since I''ve "mechanized" and I''m
running into an idiot simple
> problem.  I''m looking for more detailed examples to refresh my
memory.
> 
> My question is how to get a list of links off of a page based on
information
> in the img src attribute, specifically if the img tag contains a
> "mp3.gif/jpg".
> I, for the life of me, can''t seem to  remember how to do this.
> 
> And, as I go through this exercise, I am sure I will run into other issues
> as well.  So, if anyone has any good sources for examples, I''d
love to see
> them.
Give something like this a try:

  agent = WWW::Mechanize.new
  page  = agent.get(''http://localhost/~aaron/test.html'')
  
  links = page.search(''//a'').select { |link|
    img = link.search(''//img'').first
    img && img[''src''] =~ /mp3\.(gif|jpg)/
  }

Page can be searched like Hpricot, and links found through that interface can 
be clicked.  So feel free to get specific with your Hpricot queries.

-- 
Aaron Patterson
http://tenderlovemaking.com/