Custom recipes (archive, read-only)

[rty]

Originally Posted by Starson17

If you want help on multipage, search in this thread, then feel free to ask. Learning to do a multipage recipe is lots of fun. Print article methods are easier, but you learn less. Start here for multipage.

Thank you very much Starson17. I am looking forward to contribute more once I know how to do the Multipage recipe. Again thanks.

[taliesin1077]

Originally Posted by Starson17

If you want help on multipage, search in this thread, then feel free to ask. Learning to do a multipage recipe is lots of fun. Print article methods are easier, but you learn less. Start here for multipage.

The MaxPC articles look GREAT.

Multipage would also benefit the built-in Christian Science Monitor recipe.

[HaoleKai]

As capidamonte requested, I too would appreciate if/when someone has the time to write a recipe for the newly merged Honolulu Star-Advertiser (I'm gonna miss the Advertiser)

http://www.staradvertiser.com./rss/


Also, the Boston Globe (Boston) recipe hasn't really worked in the last month or so (since I first started using it). Sometimes it pulls a couple of stories but most of the time it pulls nothing.

Thanks!

[Daanish87]

Hi guys, I would really appreciate if I could get a custom recipe for the globe and mail business section without all the extra pictures and icons and crazy stuff... just plain text would be great:

http://www.theglobeandmail.com/repor...s/?service=rss

And the Investment executive (same without the noise, just the text):

http://www.investmentexecutive.com/f...eakingnews.xml

If someone could, I'd also like some help as to how to do these myself. I'm completely new to programming so I know basically zero. Some good resources would be appreciated.

[Starson17]

Originally Posted by Daanish87

Hi guys, I would really appreciate if I could get a custom recipe for the globe and mail business section without all the extra pictures and icons and crazy stuff... just plain text would be great:

http://www.theglobeandmail.com/repor...s/?service=rss

And the Investment executive (same without the noise, just the text):

http://www.investmentexecutive.com/f...eakingnews.xml

If someone could, I'd also like some help as to how to do these myself. I'm completely new to programming so I know basically zero. Some good resources would be appreciated.

You can start here:
http://calibre-ebook.com/user_manual...asicnewsrecipe

It looks like you already have the meat of your recipe, but want to get rid of "extra pictures and icons and crazy stuff." There are 2 ways to do it. First, you can try the print_version method. See the link, but basically it means substituting a link to a clean version for your messy version. You have to find that link yourself on the site. The other is to clean up or avoid the "crazy stuff."

I suggest you use FireFox and Firebug to identify what you want to remove or keep, then use keep_only_tags or remove_tags to either get only what you want, or remove the junk. Have fun and feel free to ask here for help.

[Daanish87]

Originally Posted by Starson17

You can start here:
http://calibre-ebook.com/user_manual...asicnewsrecipe

It looks like you already have the meat of your recipe, but want to get rid of "extra pictures and icons and crazy stuff." There are 2 ways to do it. First, you can try the print_version method. See the link, but basically it means substituting a link to a clean version for your messy version. You have to find that link yourself on the site. The other is to clean up or avoid the "crazy stuff."

I suggest you use FireFox and Firebug to identify what you want to remove or keep, then use keep_only_tags or remove_tags to either get only what you want, or remove the junk. Have fun and feel free to ask here for help.

Hi Starson, thanks for the reply. I actually wanted to try the print edition way, but here is where I got stuck. So for example, with investment executive:

The normal URL is:

http: // www .investmentexecutive.com/client/en/News/DetailNews.asp?id=53928&IdSection=146&cat=146&BIma geCI=1

The print is:

http: // www. investmentexecutive.com/client/en/News/ImprimerDetail.asp?id=53928&IdSection=146&cat=146& BImageCI=1

The URLs are nearly identical. I tried messing around with it but it keeps giving me the same output. Would you be able to guide me as to what I should put for the replacement when writing the code? Thanks

[Starson17]

Originally Posted by Daanish87

I actually wanted to try the print edition way, but here is where I got stuck. The URLs are nearly identical. I tried messing around with it but it keeps giving me the same output. Would you be able to guide me as to what I should put for the replacement when writing the code?

You want to replace DetailNews with ImprimerDetail.

Try this (No guarantees - I don't have anywhere to test it right now):

    def print_version(self, url):
        return url.replace('http://www.investmentexecutive.com/client/en/News/DetailNews', 'http://www.investmentexecutive.com/client/en/News/ImprimerDetail')

[kiklop74]

Originally Posted by HaoleKai

As capidamonte requested, I too would appreciate if/when someone has the time to write a recipe for the newly merged Honolulu Star-Advertiser (I'm gonna miss the Advertiser)

http://www.staradvertiser.com./rss/


Also, the Boston Globe (Boston) recipe hasn't really worked in the last month or so (since I first started using it). Sometimes it pulls a couple of stories but most of the time it pulls nothing.

Thanks!

The updated recipes for these will be part of the next calibre release.

[kiklop74]

Originally Posted by Daanish87

Hi guys, I would really appreciate if I could get a custom recipe for the globe and mail business section without all the extra pictures and icons and crazy stuff... just plain text would be great:

http://www.theglobeandmail.com/repor...s/?service=rss

What is wrong with the existing Globe and mail recipe?

[Daanish87]

Originally Posted by kiklop74

What is wrong with the existing Globe and mail recipe?

There are alot of pictures and other things which cause the reader to be really slow. However, I countered this by converting the .epub to pdf, and now its smooth. So maybe I dont need to really refine it much.

With Investment executive, the writing becomes condensed into one side, and even at highest zoom, the letters cannot be read.

[Daanish87]

Originally Posted by Starson17

You want to replace DetailNews with ImprimerDetail.

Try this (No guarantees - I don't have anywhere to test it right now):

    def print_version(self, url):
        return url.replace('http://www.investmentexecutive.com/client/en/News/DetailNews', 'http://www.investmentexecutive.com/client/en/News/ImprimerDetail')

Hi starson, I tried this earlier, didn't work.

[Starson17]

Originally Posted by Daanish87

Hi starson, I tried this earlier, didn't work.

Yes it did. I just tried it. It grabbed the print link you wanted ..... of course, that's not the end of the story. You still have to make the page work. Try looking at the source of the page.
All the content you want is in there.

Which road do you want to go down next - fix the non-print version or fix the print version?

[Daanish87]

Originally Posted by Starson17

Yes it did. I just tried it. It grabbed the print link you wanted ..... of course, that's not the end of the story. You still have to make the page work. Try looking at the source of the page.
All the content you want is in there.

Which road do you want to go down next - fix the non-print version or fix the print version?

Thanks Starson. Actually this is what happened. I had earlier gotten the RSS for the Globe and Mail in epub format. This was causing my reader to slow down alot and even freeze. So I converted the Globe and Mail epub to pdf and then it was working fine.

So I assumed the same thing with Investment Executive, I just converted it straight instead of checking the epub version, and the pdf was coming all strange, lots of images and microscopic writing. However I decided to give the epub version of IE a try, and lo and behold, it isn't causing my reader to slow down or freeze up, and in fact its very clear with minimal distractions(using the code you gave me).

So I guess I should take it on a case by case basis. So far, for Globe and mail, the pdf conversion seems to work and for IE the epub version seems to work.

Thanks for you help, I appreciate it.

[Daanish87]

Originally Posted by Starson17

Yes it did. I just tried it. It grabbed the print link you wanted ..... of course, that's not the end of the story. You still have to make the page work. Try looking at the source of the page.
All the content you want is in there.

Which road do you want to go down next - fix the non-print version or fix the print version?

Actually, the way it is now is fine, its not perfect but its very readable.

I do have one question. I was going through the calibre page for converting rss, and in the section on 'slicing and dicing' I was wondering where the code shown is from? And am I supposed to copy that entire code into the recipe? Or is this something else?

[Starson17]

Originally Posted by Daanish87

Actually, the way it is now is fine, its not perfect but its very readable.

I do have one question. I was going through the calibre page for converting rss, and in the section on 'slicing and dicing' I was wondering where the code shown is from? And am I supposed to copy that entire code into the recipe? Or is this something else?

I'm not really sure what you're asking. Take a look at some builtin recipes to see how the various methods are used.