Custom recipes (archive, read-only)

[capidamonte]

Originally Posted by kiklop74

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

Thanks!

cap

[HaoleKai]

Originally Posted by kiklop74

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

mahalo nui loa!
(thank you very much!)

[gambarini]

libero-news.it

Better viewing

[gambarini]

Take a look at this feed:
http://www.auto.it/rss/prove+6.xml

Every link in the feed is about the first "tablet" of the entire article; the others tablets are about others arguments regarding the same "car". The link for all tablets into the same article change only in one part, and always in the same way for every other article (about others cars).

So an example:

http://www.auto.it/prove/bmw/x6/2010/05-4850/scheda/BMW+X6+M.+Il+SUV+pi%C3%B9+veloce

http://www.auto.it/prove/bmw/x6/2010...i%C3%B9+veloce

http://www.auto.it/prove/bmw/x6/2010...i%C3%B9+veloce

http://www.auto.it/prove/bmw/x6/2010...i%C3%B9+veloce

http://www.auto.it/prove/bmw/x6/2010...i%C3%B9+veloce

http://www.auto.it/prove/bmw/x6/2010...i%C3%B9+veloce

http://www.auto.it/prove/bmw/x6/2010...i%C3%B9+veloce

is there the possibility to add the others tablets to the link of the feeds?
One option is to create an epub only for this feed, and use every link of the feed like a feed (a section title), and automatic create the details of the feed with all the links regarding tablet of the same article.

Is it correct?

[Daanish87]

Originally Posted by kiklop74

What is wrong with the existing Globe and mail recipe?

Im an idiot please ignore my posts regarding the globe and mail. I didn't even notice the built in recipes. I modified the existing one just to get my desired feeds and it works perfectly.

[Starson17]

Originally Posted by gambarini

Take a look at this feed:
http://www.auto.it/rss/prove+6.xml

I looked. First, let me fix your English. It's far, far better than my Italian, but you used the word "tablet" and it took me a while to figure out you were asking about "tabs" on the article links that the feed sent you to.

Every link in the feed links you to the first "tab" on an article page that has additional tabs; the additional tabs are about the same "car". The link for all tabs for each car on the first page/tab change only in one part, and always in the same way for every other article (about other cars).

The above revised phrasing would be a bit more clear, in case others want to pitch in. Basically, there is a single feed for multiple articles, each article being about a single car. The link takes you to a first page, and that page has additional tabs, each one being created the same way, so if you know the article link for a car, you can figure out the links to the other tabs for that car.

is there the possibility to add the others tabs to the link of the feeds?
One option is to create an epub only for this feed, and use every link of the feed like a feed (a section title), and automatic create the details of the feed with all the links regarding tabs of the same article.

Is it correct?

Yes, you could do it that way, and I think it makes the most sense. You could scrape the RSS feed as a web page with parse_index. You would build a separate feed for each car, then build the article links for each car feed from the known way that the links for the tabs are created.

Another way would be to try to just allow recursion to the deeper tab links on each page. You might have to modify the page to get access to the links. This would give you a single feed, with a link to each car, and your reader would then link from the first page to the next pages.

A third way would be to use preprocess_html and the soup to grab the tab links and add them to each page - the way multipage is done.

[gambarini]

I have created the collection of articles, but i don't understand how merge this collection to the other parsed feed.

[lordvetinari2]

Hello,

I am trying to create a new recipe for Publico PT, because the bundled one does not work any more. I tried fixing it, but there was a regexp there that I do not understand. So I tried creating my own from scratch, and it mostly works, but I have a few problems.

The main problem is that certain parts of the website work differently, for some reason. Now, my recipe is checking for certain tags to remove unnecessary items (menus, icons, etc), and not every section uses the same tags. For instance, "http://economia.publico.pt/Noticia/g...e-2013_1441633" has the main article inside a div id="content", but the article "http://desporto.publico.pt/noticia.aspx?id=1441651" has a the main article inside a div class="containerMain". As my recipe stands now, when it does not find the tag defined under "keep_only_tags", it just returns an empty page. Is there any way to say, "check for this tag and if it's not there, check for the other one instead"?

I tried using the print_version method, but another problem is that not every article from the RSS feed has the same URI structure. As you can see in the example above, sometimes they use the full title for the URI and sometimes they use a number reference. Additionally, the print ("Imprimir" in Portuguese) mechanism for article URIs such as "http://desporto.publico.pt/noticia.aspx?id=1441651", fails miserably in the browser, while the one for "http://economia.publico.pt/Noticia/g...e-2013_1441633" just prints the current URI as it is displayed, without getting rid of unnecessary formatting.

Please find attached a calibre log and the recipe as it currently stands. If an operator gives me permission, I can also upload the epub file with the daily news as retrieved with my recipe.

I hope I can get some insight on this. Thanks in advance!

[Starson17]

Originally Posted by gambarini

I have created the collection of articles, but i don't understand how merge this collection to the other parsed feed.

AFAIK, you can't (at least not easily). You either build the entire set of feeds and associated articles, with parse_index (not merge) or you use the parsed feed/articles it builds, and modify the pages it fetches (by adding/allowing links or by using the multipage type method). Those are the three methods referred to above that I can think of.

[Starson17]

Originally Posted by lordvetinari2

I am trying to create a new recipe for Publico PT, because the bundled one does not work any more. I tried fixing it, but there was a regexp there that I do not understand.

It might be easier to just figure out this regex.

So I tried creating my own from scratch, and it mostly works, but I have a few problems.

The main problem is that certain parts of the website work differently, for some reason. Now, my recipe is checking for certain tags to remove unnecessary items (menus, icons, etc), and not every section uses the same tags. For instance, "http://economia.publico.pt/Noticia/g...e-2013_1441633" has the main article inside a div id="content", but the article "http://desporto.publico.pt/noticia.aspx?id=1441651" has a the main article inside a div class="containerMain". As my recipe stands now, when it does not find the tag defined under "keep_only_tags", it just returns an empty page. Is there any way to say, "check for this tag and if it's not there, check for the other one instead"?

Yes - sort of, but why not just keep both tags instead of making it conditional? It should work just as well, unless both tags are used on the same page.

[gambarini]

Originally Posted by Starson17

AFAIK, you can't (at least not easily). You either build the entire set of feeds and associated articles, with parse_index (not merge) or you use the parsed feed/articles it builds, and modify the pages it fetches (by adding/allowing links or by using the multipage type method). Those are the three methods referred to above that I can think of.

Yes, my eng isn't so good...
I can print the article tuple in debug, and the feed that i have created seems correct,
so these are my next steps:
1°) create the epub with only this feed
2°) create all other feeds (it could be only a simple "append" of the standard tag "title", "link", "description", "date" and so on), using the same "parse_index" used for the step one, for add (not merge ) all others feed.

It is also a good solution, i think, for my previous (and not yet solved) problem with the "la_stampa" recipe (There is two or three feeds without the title for every article).

[Starson17]

Originally Posted by gambarini

Yes, my eng isn't so good...

It's better than my Italian

I can print the article tuple in debug, and the feed that i have created seems correct,
so these are my next steps:
1°) create the epub with only this feed
2°) create all other feeds (it could be only a simple "append" of the standard tag "title", "link", "description", "date" and so on), using the same "parse_index" used for the step one, for add (not merge ) all others feed.

It is also a good solution, i think, for my previous (and not yet solved) problem with the "la_stampa" recipe (There is two or three feeds without the title for every article).

Yes.

[gambarini]

Recipe for
corrieredellosport (sport daily newspaper)
auto & autosprint (formula 1 and car news)

[flyash]

I am using Calibre to fetch updates to the following blog: http://www.bodyrecomposition.com/feed/rss.

But the width of the resulting ebook is greater than my Sony 600 screen, so the right edge of the page is chopped off. Is there a way to modify the recipe to correct this?

[kovidgoyal]

Use

conversion_options = {'linearize_tables':True}