Custom recipes (archive, read-only)

[kiklop74]

The Wired Magazine (do not confuse with the existing wired recipe this one downloads entire magazine content available on the web).

[Over]

Originally Posted by kiklop74

New recipe for ionline.pt:

Thank you very much! It's working great.

Can I ask one last favour?

I would really be thankful for a recipe for:

http://www.nursingtimes.net/rss-feeds/

I've tried the basic way, but it doesn't work.

[kiklop74]

New recipe for neowin.net:

[tnronin]

Originally Posted by dwanthny

Unzip the the files into a directory.

Open Calibre
Drop down arrow next to Fetch news
Add custom news source
Load recipe from file (select nationalreviewonline.recipe)

I just followed these steps to verify things.

Good Luck

&

Awesome! It is working now.

[Krittika Goyal]

@jayman
let me know if you get the same epub file

[XanthanGum]

Kovid,

I'm attaching an improved recipe for Common Dreams. I hope you can replace the current version in Calibre.

Thanks...

XG

[kovidgoyal]

Originally Posted by XanthanGum

Kovid,

I'm attaching an improved recipe for Common Dreams. I hope you can replace the current version in Calibre.

Thanks...

XG

Sure .

[andersent]

Hi All,

Shouldn't I be able to go to a website that has a table of contents of web links for an online book and essentially populate the feeds [] = section of a recipe by looping through the html links on the TOC page? This way, if you find a site that has a book or a bunch of links to content that you would like on your eReader, you don't have to point to every link by name on a page if there are a good deal of links. Below is a concrete example that will help clarify what I am looking to do:

I could create my own feed.xml file and put it on a webpage somewhere and point calibre at this feed. The feed would simply specify the following link (which has a table of contents with links on the same site to the book): http://www.newmanreader.org/works/idea/index.html

Now, if I point calibre to just this page, it will just see all the links but to my knowledge (unless I'm missing something) it can't follow every link on this page. So, rather than specifying each link that is on this page in the feeds.xml file, can't I just have calibre read the html, look for ahref= tags on this page, and then dynamically build a list of feeds[] = based on the ahref tags? It seems like some of the newspaper recipes essentially do this but I can't always follow the logic of what is happening in the pre built recipes. I think this would make custom recipes and calibre very powerul, as you could go to a website that has a listing of a mess of links and have it go download all of the content and section it into "articles" on your eReader, but within one "book."

[kiklop74]

Originally Posted by andersent

Hi All,

Shouldn't I be able to go to a website that has a table of contents of web links for an online book and essentially populate the feeds [] = section of a recipe by looping through the html links on the TOC page?

You are able to do that. In fact all recipes do exactly that. The difference is the source of the links. It is either RSS feed (which is just an xml), or common html page.

In case of RSS feeds everything is more or less automated, custom programming is required for extracting links from common web page.

For an example of this just take a look at this page:

http://calibre-ebook.com/user_manual/news.html

[DoctorOhh]

I downloaded news, while d/ling an image was not found 404 error. This resulted in this image string being in the final epub.

<img src="http://wattsupwiththat.com/Users/Anthony/AppData/Local/Temp/moz-screenshot-3.png" alt="" class="calibre9"/>

Instead of something like this

<img src="../../feed_1/article_10/images/img3.jpg.jpg" alt="" class="calibre9"/>

When my PRS-505 tried to access this image it locked up and eventually restarted. I don't have the experience or knowledge required to get rid of this in post processing. Is there any way to clean up the code prior to creating the epub that would eliminate any images that have <img src="http to start the string?

I know how to eliminate all images but I find the images valuable. I just want to remove the images that couldn't be downloaded and thus end up in the epub calling an image from a url.

[kovidgoyal]

def postprocess_html(self, soup, first):
   for img in soup.findAll('img', src=True):
       if img['src'].startswith('http:'): img.extract()

[DoctorOhh]

Originally Posted by kovidgoyal

def postprocess_html(self, soup, first):
   for img in soup.findAll('img', src=True):
       if img['src'].startswith('http:'): img.extract()

Thank you I'll give it a go.

Update: I appended this post process onto the offending site recipe, but they corrected their page so I have to assume it works great. One thing for sure it isn't doing any harm.

Thanks again.

[kovidgoyal]

I forgot that the last line in postprocess_html should be

return soup

[Krittika Goyal]

@jayman
let me know if this one is ok?

[DoctorOhh]

Originally Posted by kovidgoyal

I forgot that the last line in postprocess_html should be

return soup

I got that from many of the implementations.

Thanks again for helping fill in the gaps.