Atom & the Standard Library RSS Module

ruby-talk
1

I need to parse RSS 1.0, 2.0 and ATOM feeds. I upgraded my RSS module
to the latest version (0.2.4) to get ATOM support. The strange thing
is that the data structure returned for ATOM feeds is ugly and wildly
inconsistent with the nice, clean one that is returned for RSS feeds.

I've noticed there are a couple of competing Ruby ports of Mark
Pilgrim's Universal Feedparser. The 'rfeedparser' one looks to be the
best and FeedTools looks interesting, but I haven't actually tried
them yet. (I really like the Universal Feedparser for Python.)

Does anyone have any suggestions on which direction to take?

2

Hi,

In <0ba46612-3b54-495d-81a2-2c05006c3b21@k1g2000prb.googlegroups.com>
  "Atom & the Standard Library RSS Module" on Sat, 5 Apr 2008 06:25:05 +0900,

···

grant <gaustin@gmail.com> wrote:

I need to parse RSS 1.0, 2.0 and ATOM feeds. I upgraded my RSS module
to the latest version (0.2.4) to get ATOM support. The strange thing
is that the data structure returned for ATOM feeds is ugly and wildly
inconsistent with the nice, clean one that is returned for RSS feeds.

Please show an example.

--
kou

3

Sorry, I got tangled up in my own wishes. My problem with the library
was just in my head. I suppose I was just hoping for a more consistent
representation of a feed than I've had to deal with in the past. I
know that what is generated by RSS/ATOM parsing libraries is a
reflection of the feeds themselves.

Anyway, a few examples of what bugs me, demonstrated in irb:

require 'rss'

rss = 'http://www.giftedslacker.com/feed/&#39;
atom = 'http://oblivionation.blogspot.com/feeds/posts/default&#39;

rssfeed = RSS::Parser.parse(rss)
atomfeed = RSS::Parser.parse(atom)

#print the content of the most recent post
puts rssfeed.items[0].content_encoded
puts atomfeed.items[0].content.content

#print the titles of the posts in the feed
rssfeed.items.each {|item| puts item.title}
atomfeed.items.each {|item| puts item.title.content}

#print the author of the most recent post
rssfeed.items[0].dc_creator
atomfeed.entries[0].author.name.content

···

On 5 Apr, 01:29, Kouhei Sutou <k...@cozmixng.org> wrote:

Please show an example.

---

What I'd like is a consistent interface to the commonly used elements
in ATOM and RSS feeds, regardless of version. It must be more
difficult than it seems to me, because I'm not aware that anyone does
it. I might give it a try just for fun. But, being new to Ruby, I'm
not sure where to begin.

-grant

4

grant wrote:

What I'd like is a consistent interface to the commonly used elements
in ATOM and RSS feeds, regardless of version. It must be more
difficult than it seems to me, because I'm not aware that anyone does
it. I might give it a try just for fun. But, being new to Ruby, I'm
not sure where to begin.

You could build upon/look at/submit patches to Simple RSS:
http://simple-rss.rubyforge.org/

I've used it, and while it is simple to use, it comes at the cost of
limited functionality (as far as I could see. I only used it to grab
NetBeans Ruby IDE builds off the web, when the buildserver used updated
its RSS feed, so take my comments with a grain of salt.)

- -- Phillip Gawlowski

···

On 5 Apr, 01:29, Kouhei Sutou <k...@cozmixng.org> wrote:

5

Hi,

In <8311b2d9-35ea-4ff9-a294-8872eb52dbfa@c65g2000hsa.googlegroups.com>
  "Re: Atom & the Standard Library RSS Module" on Sat, 5 Apr 2008 20:45:06 +0900,

> Please show an example.

Sorry, I got tangled up in my own wishes. My problem with the library
was just in my head. I suppose I was just hoping for a more consistent
representation of a feed than I've had to deal with in the past. I
know that what is generated by RSS/ATOM parsing libraries is a
reflection of the feeds themselves.

Anyway, a few examples of what bugs me, demonstrated in irb:

require 'rss'

rss = 'http://www.giftedslacker.com/feed/&#39;
atom = 'http://oblivionation.blogspot.com/feeds/posts/default&#39;

rssfeed = RSS::Parser.parse(rss)
atomfeed = RSS::Parser.parse(atom)

You can normalize parsed feed by to_rss, to_atom or
to_xml. For example:

  rss10feed = atomfeed.to_rss("1.0")

You may need to set default value:

  rss10feed = atomfeed.to_rss("1.0") do |maker|
    maker.channel.about ||= maker.channel.link
    maker.channel.description ||= "No description"
    maker.items.each do |item|
      item.title ||= "UNKNOWN"
      item.link ||= "UNKNOWN"
    end
  end

#print the content of the most recent post
puts rssfeed.items[0].content_encoded
puts atomfeed.items[0].content.content

#print the titles of the posts in the feed
rssfeed.items.each {|item| puts item.title}
atomfeed.items.each {|item| puts item.title.content}

#print the author of the most recent post
rssfeed.items[0].dc_creator
atomfeed.entries[0].author.name.content

The atom specification says that all atom element may have
xml:base and xml:lang attributes. If
RSS::atom::Entry::Author#name returns a String, we can't
get such attribute values. This is why
RSS::atom::Entry::Author#name returns an
RSS::atom::Entry::Author::Name not a String.

BTW, what about the following API?

  atomfeed.entries[0].author.name # => a String
  atomfeed.entries[0].author.name do |name|
                                    # name: a RSS::atom::Entry::Author::Name
    name.content # => a String
  end # => the last evaluated value (name.content)

Thanks,

···

grant <gaustin@gmail.com> wrote:

On 5 Apr, 01:29, Kouhei Sutou <k...@cozmixng.org> wrote:

--
kou