Custom news fetch: Unreadable text

[dumanb]

Hi guys,

I am kind of new in the forum. Could you help me with this please?

I created a custom receipt with Calibre and added some feeds. Calibre downloads the feeds and sends them to my kindle but the text of some feeds are not readable; they look like bunch of codes like this: ÐZ¸«ðÌÞOo—Lf#!)vŸÂÕDzÃIï¢$è8é a…zŸf#¿…üéôë˜úÜÉ¡ÙMèÊhƒ=¤¢oLýÃYétº°~·âÆ/‰]'òÿŠ ½ð�ÖçâêøҮ߬üΡÉèšJ�åPÉeüŽTäù°z


What's that I am doing wrong? Did I used the wrong feed url?
This is how does it looks like when I click the advanced mode:

#!/usr/bin/env python2
# vim:fileencoding=utf-8
from __future__ import unicode_literals, division, absolute_import, print_function
from calibre.web.feeds.news import BasicNewsRecipe

class AdvancedUserRecipe1453356016(BasicNewsRecipe):
title = 'Feeds'
oldest_article = 1
max_articles_per_feed = 100
auto_cleanup = True

feeds = [
('Experimental Economics', 'http://nep.repec.org/rss/nep-exp.rss.xml'),
('Development Economics', 'http://nep.repec.org/rss/nep-dev.rss.xml'),
('Experimental Economics', 'http://link.springer.com/search.rss?facet-content-type=Article&facet-journal-id=10683&channel-name=Experimental%20Economics'),
('Evolutionary Economics', 'http://nep.repec.org/rss/nep-evo.rss.xml'),
('IZA Labor Economics', 'http://www.izajole.com/latest/rss'),
('Labour Economics', 'http://nep.repec.org/rss/nep-lab.rss.xml'),
('Social Norms And Social Capital', 'http://nep.repec.org/rss/nep-soc.rss.xml'),
('The Quarterly Journal Of Economics', 'http://qje.oxfordjournals.org/rss/current.xml'),
('Evonomics', 'http://evonomics.com/feed/'),
('Cultural Economics', 'http://nep.repec.org/rss/nep-cul.rss.xml'),
('Law and Economics', 'http://nep.repec.org/rss/nep-law.rss.xml'),
('Public Economics', 'http://nep.repec.org/rss/nep-pbe.rss.xml'),
('Positive Political Economics', 'http://nep.repec.org/rss/nep-pol.rss.xml'),
]

[eschwartz]

You will notice that each of those articles have a source link -- which points to a PDF file.

Note for the future: calibre's recipe subsystem really doesn't like it when you try to pretend PDF files are actually HTML files.

[dumanb]

Thanks for explaining.
But is there any solution? Can I enter an address that directly gets the pdf and converts it to a kindle version?

[eschwartz]

I have no idea -- I don't really use the recipes feature.

You could certainly download the PDF and manually add it as a new book in calibre, but I am afraid I have no idea how to scrape for PDFs using a recipe -- if it is even possible.

[Divingduck]

Take a look on re-usable code post #16. achims made a nice snipped of code you maybe can use for your own recipe exercises. You will find a usage for the code in an old recipe: https://www.mobileread.com/forums/sho...d.php?t=155792

[dumanb]

This is to complicated for me.

If the feed url ends with .xml which usually is the case, calibre won't convert it to kindle version (.mobi), meaning that Calibre Fetch news is useless for kindle.

It would be surprising if no one came up with a solution yet.

[PeterT]

The calibre news feed is really designed to handle regular web sites / feeds; one where the links are to normal HTML type pages. The site you are scripting contains material in PDF format. The fact that links end in XML has nothing to do with the inability to view on a Kindle; this will not work on an ePub reader either.

[eschwartz]

Originally Posted by dumanb

This is to complicated for me.

If the feed url ends with .xml which usually is the case, calibre won't convert it to kindle version (.mobi), meaning that Calibre Fetch news is useless for kindle.

It would be surprising if no one came up with a solution yet.

The feed is .xml because it is XML.

Feeds are full of links -- the purpose of a feed is to tell you where the REAL articles are.

Unfortunately, the articles are provided as .pdf , which you will need to download separately since Fetch News is not designed to work with PDF (it does not matter what ereader you have, it still won't download in the first place).