Custom recipes (archive, read-only)

[DoctorOhh]

Originally Posted by Dereks

Hi guys,
I was searching quite a lot through the topic, but didn't manage to find the exact solution.
I need to have a google reader recipe, which will pick up only starred items from the feed (no need for tagging and or storing it all in sub folders, just plain simple starred items). If I understand correctly that's how default google recipe used to work.

Have you tried the Google Reader Recipe that comes with Calibre? It is under Unknown not a country. Let us know if it works.

[Dereks]

yes, I did.
It creates an empty page under "starred" and "broadcast" titles and then pick up all my tags and folders from the google reader.

[Daephex]

Would it be possible to build a sort of "dragnet" recipe that would consistently search blogs for a certain phrase, and return me the results once a day? I have no programming skills whatsoever, but I wonder if this couldn't be done through Technorati or even Google's blog search somehow.

[p4np5n]

Hi,

I am trying to create a recipe for a greek news website. The recipe so far looks like that:

from calibre.web.feeds.recipes import BasicNewsRecipe

class Kathimerini(BasicNewsRecipe):
    title                  = 'Kathimerini'
    __author__             = 'Pan'
    description            = 'News from Greece'
    oldest_article         = 2
    max_articles_per_feed  = 100
    encoding               = 'windows-1253'
    publisher              = 'Kathimerini'
    category               = 'news, GR'
    language               = 'el'

    feeds          = [
                      ('Πολιτική', 'http://wk.kathimerini.gr/xml_files/politics.xml'),
                    ]

    def print_version(self, url):
        return url.replace('http://news.kathimerini.gr/4dcgi/', 'http://news.kathimerini.gr/4dcgi/4dcgi/')

The feed I am trying it on is this:
http://wk.kathimerini.gr/xml_files/politics.xml

To get a printable view from a news item, say:
http://news.kathimerini.gr/4Dcgi/_w_...02/2010_390924

one has to add another "4Dcgi" string in the url:
http://news.kathimerini.gr/4Dcgi/4Dc...02/2010_390924
(I believe I've done it correctly. )

However, when executing the recipe, it seems to be unable to get any articles from the feed. any ideas?

[p4np5n]

Ok, I managed to find out what's wrong with the above. I would therefore like to submit (the first) two greek recipes for calibre!

Kathimerini

Spoiler:

from calibre.web.feeds.recipes import BasicNewsRecipe

class Kathimerini(BasicNewsRecipe):
title = 'Kathimerini'
__author__ = 'Pan'
description = 'News from Greece'
max_articles_per_feed = 100
oldest_article = 100
publisher = 'Kathimerini'
category = 'news, GR'
language = 'el'
no_stylesheets = True
remove_tags_before = dict(name='td',attrs={'class':'news'})
remove_tags_after = dict(name='td',attrs={'class':'news'})
remove_attributes = ['width', 'src','header','footer']

feeds = [
('Πολιτική', 'http://wk.kathimerini.gr/xml_files/politics.xml'),
('Ελλάδα', ' http://wk.kathimerini.gr/xml_files/ell.xml'),
('Κόσμος', ' http://wk.kathimerini.gr/xml_files/world.xml'),
('Οικονομία', 'http://wk.kathimerini.gr/xml_files/economy_1.xml'),
('Επιχειρήσεις', 'http://wk.kathimerini.gr/xml_files/economy_2.xml'),
('Διεθνής Οικονομία', 'http://wk.kathimerini.gr/xml_files/economy_3.xml'),
('Πολιτισμός', 'http://wk.kathimerini.gr/xml_files/civ.xml'),
('Μόνιμες Στήλες', 'http://wk.kathimerini.gr/xml_files/st.xml'),
]

def print_version(self, url):
return url.replace('http://news.kathimerini.gr/4dcgi/', 'http://news.kathimerini.gr/4dcgi/4dcgi/')

and Ta Nea

Spoiler:

class TaNea(BasicNewsRecipe):
title = u'Ta Nea'
__author__ = 'Pan'
oldest_article = 1
max_articles_per_feed = 100
no_stylesheets = True

remove_tags_before = dict(name='div',attrs={'id':'print-body'})
remove_tags_after = dict(name='div',attrs={'id':'text'})
feeds = [
(u'Ελλάδα', u'http://www.tanea.gr/default.asp?pid=66&la=1'),
(u'Κόσμος', u'http://www.tanea.gr/default.asp?pid=67&la=1'),
(u'Οικονομία', u'http://www.tanea.gr/default.asp?pid=68&la=1'),
(u'Πολιτισμός', u'http://www.tanea.gr/default.asp?pid=69&la=1'),
(u'Γνώμες', u'http://www.tanea.gr/default.asp?pid=79&la=1'),
(u'Ριπές', u'http://www.tanea.gr/default.asp?pid=80&la=1'),
(u'Αιχμές', u'http://www.tanea.gr/default.asp?pid=81&la=1')
]

def print_version(self, url):
return url.replace('http://www.tanea.gr/default.asp?pid=2', 'http://www.tanea.gr/default.asp?pid=96')

[kovidgoyal]

This would be very inefficient to build in calibre as calibre would have to go an search every blog page, every time. Instead if there is some web service that does this, say google alerts, that also provides an RSS feed, it may be possible to use that to build a recipe.

Originally Posted by Daephex

Would it be possible to build a sort of "dragnet" recipe that would consistently search blogs for a certain phrase, and return me the results once a day? I have no programming skills whatsoever, but I wonder if this couldn't be done through Technorati or even Google's blog search somehow.

[rylsfan]

I am trying to work in a feed from the Kansas City Star. When I go to print version they embed a v-print in the middle.

http://www.kansascity.com/sports/roy...y/1750645.html

http://www.kansascity.com/sports/roy...y/1750645.html

I was hoping this code would fix it:

def print_version(self, url):
parts=url.split('/')
id = parts[4].split('.')
if not id[0].isdigit():
id = parts[5].split('.')
return url.replace(url, 'http://www.kansascity.com/sports/royals/v-print/story/' + id[0])

But instead, I just get a page of nothing :-)

[kiklop74]

New recipe for Wired Magazine - UK edition:

Just to clarify this is a recipe for a monthly edition of Wired Magazine - not the daily feed stuff.

Update: Modified the recipe to fix cover url error

[kiklop74]

Originally Posted by rylsfan

I am trying to work in a feed from the Kansas City Star. When I go to print version they embed a v-print in the middle.

But instead, I just get a page of nothing :-)

Try this:

    def print_version(self, url):
        art, sep, rest = url.rpartition('/story/')
        return art + '/v-print/story/' + rest

[rylsfan]

That did it! All I need to do now is remove the printer icon and the site logo. Hopefully I can do that.

(knocks on wood)

[Chi]

Hi
Have been trying to make a recipe by myself, but didn't quite understand it.

I really want a recipe that's combined of these two feeds:

• ScienceDaily: Spider and Tick News http://www.sciencedaily.com/rss/plan...ls/spiders.xml
• ScienceDaily: Insect (and Butterfly) News http://www.sciencedaily.com/rss/plan...utterflies.xml

Is that possible?

Edit: I tried this, but it would not download.

class AdvancedUserRecipe1266438034(BasicNewsRecipe):
    title          = u'ScienceDaily - Arachnids and Insects'
    oldest_article = 40
    max_articles_per_feed = 100

    feeds          = [(u'ScienceDaily: Spider and Tick News', u'http://www.sciencedaily.com/rss/plants_animals/spiders.xml'), (u'ScienceDaily: Insect (and Butterfly) News', u'http://www.sciencedaily.com/rss/plants_animals/insects_and_butterflies.xml')]

[kiklop74]

There is already available Science Daily recipe. You can just edit that one to fit your needs - which means add your feeds.

You will find the recipe in {calibre inst folder}/resources/recipes/sciencedaily.recipe

[Chi]

Originally Posted by kiklop74

There is already available Science Daily recipe. You can just edit that one to fit your needs - which means add your feeds.

You will find the recipe in {calibre inst folder}/resources/recipes/sciencedaily.recipe

Thanks
But I can't seem to find the find the location or the file named "sciencedaily.recipe". I'm using Ubuntu 9.10 so maybe that's the problem?

Edit: Forget it, I found it under ./usr/lib/calibre/calibre/web/feeds/recipes

[kiklop74]

Ah so packaging is different on linux

[Dereks]

what about google recipe. If i didn't make myself clear: i am talking about default calibrie-packaged recipe. In my cases, it picks up all the stuff from folders, omits starred section.
How can it be modified to do exactly the opposite: i need only starred items.
thanks.