Recipe for Reading Blog Feeds possible ?

[markwaters]

Is this possible ?

A recipe that picks up a RSS feed containing items that include links to longer articles.
It could be used to watch a blog or twitter feed that contains links to longer articles for reading later.
I am thinking of sites / feeds like this :-

• http://thebrowser.com/best
• http://longform.org/
• Dave Winer's Link Blog

Any help would be appreciated.
Thanks.

[Starson17]

Originally Posted by markwaters

Is this possible ?

A recipe that picks up a RSS feed containing items that include links to longer articles.

All RSS feeds contain items that include links to longer articles

It could be used to watch a blog or twitter feed that contains links to longer articles for reading later.
I am thinking of sites / feeds like this :-

If those aren't RSS feeds, use parse_index.
http://manual.calibre-ebook.com/news.html#id8

[markwaters]

Normally a RSS feed contains a number of articles which contain the title and the body of the text.

RSS - Article - Title - Body

I am looking for a solution where the RSS feed contains a number of articles which contain the title but inside the body of text is a URL to the actual article I want to read.

RSS - Article - Title - Body - Link - Article - Title - Body

https://secure.wikimedia.org/wikipedia/en/wiki/Linklog

If I use my current (working) RSS feed recipe I will end up with an ebook containing links to articles.
I am hoping to have calibre go to the next link and pull that article in instead.

Does that make more sense ?

[Starson17]

Originally Posted by markwaters

Normally a RSS feed contains a number of articles which contain the title and the body of the text.

There are various RSS feed formats. Some have the body in the feed, some have only a short summary and a link to the body. Calibre's recipe system can handle all. By default, if the body is longer than 500 characters, Calibre will assume the body is in the feed, and if it's shorter, Calibre will assume the link has the article and retrieve the link. You can force either behavior.

You are asking about non-RSS feeds, aka a "linklog" or just a page that has links. Calibre can handle those as well. As I posted, parse_index is the tool for that job.

If I use my current (working) RSS feed recipe I will end up with an ebook containing links to articles.

I suspect that's because it's not an RSS feed, even though it may sort of look like one.

I am hoping to have calibre go to the next link and pull that article in instead.
Does that make more sense ?

It made sense the first time. Use parse_index and see the NYT example. You have to craft code to grab the links and create a virtual feed inside the recipe. That's what parse_index does - you tell it to read the page, how to find the links you want, what titles and summaries and dates to use and it builds the virtual RSS feed then proceeds as though there was a real RSS feed.

[Starson17]

BTW, although parse_index is the preferred way to deal with this, you can also try using "recursions" which sets the depth that links will be followed. If your pages are clean, with only links to pages you want, this might do the job.

Also, I should warn you that linklogs where the links are to random pages of interest are poor fodder for recipes. Recipes need some consistency to the pages retrieved if it is going to remove ads, apply formatting, etc. If the links point to just any old random pages, and not to pages with common attributes, the recipe results will be poor.

[markwaters]

OK , thanks for the answers , at least I know its possible and can give it a try.

[BRGriff]

Try these links for an RSS feed and go from there. It is much simpler and a recipe can be created with little difficulty.

http://feeds.thebrowser.com/feed/botm
http://longform.org/feed/
http://static.scripting.com/myReallySimple/linkblog.xml

Hope this helps. If so, please do not forget to add to my "Karma".