Feedbag

Forked version of feedbag that returns title, description and url.

Feedbag is a feed auto-discovery Ruby library. You don't need to know more about it. It is said to be:

Ruby's favorite auto-discovery tool/library!

>> require "rubygems"
=> true
>> require "feedbag"
=> true
>> Feedbag.find "log.damog.net"
=> ["http://feeds.feedburner.com/TeoremaDelCerdoInfinito", "http://log.damog.net/comments/feed/"]
>> Feedbag.feed?("google.com")
=> false
>> Feedbag.feed?("http://planet.debian.org/rss20.xml")
=> true

$ sudo gem install damog-feedbag -s http://gems.github.com/

Or just grab feedbag.rb and use it on your own project:

$ wget http://github.com/damog/feedbag/raw/master/lib/feedbag.rb

So you want to know more about it.

OK, if the URL passed to the find method is a feed itself, that only feed URL will be returned.

>> Feedbag.find "github.com/damog.atom"
=> ["http://github.com/damog.atom"]
>> 

Otherwise, it will always return LINK feeds first, A (anchor tags) feeds later. Between A feeds, the ones hosted on the same URL's host, will have larger priority:

>> Feedbag.find "http://ve.planetalinux.org"
=> ["http://feedproxy.google.com/PlanetaLinuxVenezuela", "http://rendergraf.wordpress.com/feed/", "http://rootweiller.wordpress.com/feed/", "http://skatox.com/blog/feed/", "http://kodegeek.com/atom.xml", "http://blog.0x29.com.ve/?feed=rss2&cat=8"]
>> 

On your application you should only take the very first element of the array, most of the times:

>> Feedbag.find("planet.debian.org").first(3)
=> ["http://planet.debian.org/rss10.xml", "http://planet.debian.org/rss20.xml", "http://planet.debian.org/atom.xml"]
>> 

(Try running that same example without the "first" method. That example's host is a blog aggregator, so it has hundreds of feed URLs:)

>> Feedbag.find("planet.debian.org").size
=> 104
>> 

Feedbag will find them all, but it will return the most important ones on the first elements on the array returned.

>> Feedbag.find("cnn.com")
=> ["http://rss.cnn.com/rss/cnn_topstories.rss", "http://rss.cnn.com/rss/cnn_latest.rss", "http://rss.cnn.com/services/podcasting/robinmeade/rss.xml"]
>> 

• Because it's cool.
• Because it only uses Hpricot as dependency.
• Because it follows modern feed filename conventions (like those ones used by WordPress blogs, or Blogger, etc).
• Because it's a single file you can embed easily in your application.
• Because it passes most of the Mark Pilgrim's Atom auto-discovery test suite. It doesn't pass them all because some of those tests are broken (citation needed).

• Because I liked Benjamin Trott's Feed::Find.
• Because I thought it would be good to have Feed::Find's functionality in Ruby.
• Because I thought it was going to be easy to maintain.
• Because I was going to use it on rFeed.
• And finally, because I didn't know rfeedfinder existed :-)

Please, report bugs to rt@support.axiombox.com or directly to the author.

git clone git://github.com/damog/feedbag.git

...patch, build, hack and make pull requests. I'll be glad.

David Moreno <david@axiombox.com>.

This is free software. See COPYING for more information.

Raquel, for making Axiombox and most of my dreams possible. Also, GitHub for making a nice code sharing service that doesn't suck.