Cutlery/calibre
1
0
Fork 0
mirror of https://github.com/kovidgoyal/calibre.git synced 2025-07-09 03:04:10 -04:00
Code Issues Packages Projects Releases Wiki Activity

Update Dilema Veche

Browse Source
This commit is contained in:
Kovid Goyal 2013-09-10 14:34:49 +05:30
parent c43a3e7969
commit c674ca1342
1 changed files with 79 additions and 38 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

117
recipes/dilemaveche.recipe
View File

@ -6,46 +6,87 @@ __copyright__ = u'2011, Silviu Cotoar\u0103'
''' '''
dilemaveche.ro dilemaveche.ro
''' '''
from calibre.web.feeds.news import BasicNewsRecipe from calibre.web.feeds.news import BasicNewsRecipe
class DilemaVeche(BasicNewsRecipe): class DilemaVeche(BasicNewsRecipe):
title = u'Dilema Veche' # apare vinerea, mai pe dupa-masa,depinde de Luiza cred (care se semneaza ca fiind creatorul fiecarui articol in feed-ul RSS)
__author__ = u'Silviu Cotoar\u0103' title = u'Dilema Veche'
description = 'Sint vechi, domnule! (I.L. Caragiale)' __author__ = 'song2' # inspirat din scriptul pentru Le Monde. Inspired from the Le Monde script
publisher = u'Adev\u0103rul Holding' description = '"Sint vechi, domnule!" (I.L. Caragiale)'
oldest_article = 5 publisher = 'Adevarul Holding'
language = 'ro' oldest_article = 7
max_articles_per_feed = 100 max_articles_per_feed = 200
no_stylesheets = True encoding = 'utf8'
use_embedded_content = False language = 'ro'
category = 'Ziare' masthead_url = 'http://www.dilemaveche.ro/sites/all/themes/dilema/theme/dilema_two/layouter/dilema_two_homepage/logo.png'
encoding = 'utf-8' publication_type = 'magazine'
cover_url = 'http://dilemaveche.ro/sites/all/themes/dilema/theme/dilema_two/layouter/dilema_two_homepage/logo.png' feeds = [
('Editoriale si opinii - Situatiunea', 'http://www.dilemaveche.ro/taxonomy/term/37/0/feed'),
conversion_options = { ('Editoriale si opinii - Pe ce lume traim', 'http://www.dilemaveche.ro/taxonomy/term/38/0/feed'),
'comments' : description ('Editoriale si opinii - Bordeie si obiceie', 'http://www.dilemaveche.ro/taxonomy/term/44/0/feed'),
,'tags' : category ('Editoriale si opinii - Talc Show', 'http://www.dilemaveche.ro/taxonomy/term/44/0/feed'),
,'language' : language ('Tema saptamanii', 'http://www.dilemaveche.ro/taxonomy/term/19/0/feed'),
,'publisher' : publisher ('La zi in cultura - Dilema va recomanda', 'http://www.dilemaveche.ro/taxonomy/term/58/0/feed'),
} ('La zi in cultura - Carte', 'http://www.dilemaveche.ro/taxonomy/term/14/0/feed'),
('La zi in cultura - Film', 'http://www.dilemaveche.ro/taxonomy/term/13/0/feed'),
keep_only_tags = [ ('La zi in cultura - Muzica', 'http://www.dilemaveche.ro/taxonomy/term/1341/0/feed'),
dict(name='div', attrs={'class':'c_left_column'}) ('La zi in cultura - Arte performative', 'http://www.dilemaveche.ro/taxonomy/term/1342/0/feed'),
] ('La zi in cultura - Arte vizuale', 'http://www.dilemaveche.ro/taxonomy/term/1512/0/feed'),
('Societate - Ieri cu vedere spre azi', 'http://www.dilemaveche.ro/taxonomy/term/15/0/feed'),
remove_tags = [ ('Societate - Din polul opus', 'http://www.dilemaveche.ro/taxonomy/term/41/0/feed'),
dict(name='div', attrs={'id':['adshop_widget_428x60']}) , ('Societate - Mass comedia', 'http://www.dilemaveche.ro/taxonomy/term/43/0/feed'),
dict(name='div', attrs={'id':['gallery']}) ('Societate - La singular si la plural', 'http://www.dilemaveche.ro/taxonomy/term/42/0/feed'),
] ('Oameni si idei - Educatie', 'http://www.dilemaveche.ro/taxonomy/term/46/0/feed'),
('Oameni si idei - Polemici si dezbateri', 'http://www.dilemaveche.ro/taxonomy/term/48/0/feed'),
remove_tags_after = [ ('Oameni si idei - Stiinta si tehnologie', 'http://www.dilemaveche.ro/taxonomy/term/46/0/feed'),
dict(name='div', attrs={'id':['adshop_widget_428x60']}) ('Dileme on-line', 'http://www.dilemaveche.ro/taxonomy/term/005/0/feed')
]
feeds = [
(u'Feeds', u'http://dilemaveche.ro/rss.xml')
] ]
remove_tags_before = dict(name='div',attrs={'class':'spacer_10'})
remove_tags = [
dict(name='div', attrs={'class':'art_related_left'}),
dict(name='div', attrs={'class':'controale'}),
dict(name='div', attrs={'class':'simple_overlay'}),
]
remove_tags_after = [dict(id='facebookLike')]
remove_javascript = True
no_stylesheets = True
remove_empty_feeds = True
extra_css = """
body{font-family: Georgia,Times,serif }
img{margin-bottom: 0.4em; display:block}
"""
needs_subscription = 'optional'
cover_margins = (10, 15, '#ffffff')
def get_browser(self):
br = BasicNewsRecipe.get_browser(self)
if self.username is not None and self.password is not None:
br.open('http://dilemaveche.ro/user/login')
br.select_form(nr=0)
br['username'] = self.username
br['password'] = self.password
br.submit()
return br
def get_cover_url(self):
cover_url = None
soup = self.index_to_soup('http://dilemaveche.ro')
link_item = soup.find('div',attrs={'class':'box_dr_pdf_picture'})
if link_item and link_item.a:
cover_url = link_item.a['href']
br = BasicNewsRecipe.get_browser()
try:
br.open(cover_url)
except: # daca nu gaseste pdf-ul
self.log("\nPDF indisponibil")
link_item = soup.find('div',attrs={'class':'box_dr_pdf_picture'})
if link_item and link_item.img:
cover_url = link_item.img['src']
br = BasicNewsRecipe.get_browser()
try:
br.open(cover_url)
except: # daca nu gaseste nici imaginea mica mica
print('Mama lor de nenorociti! nu este nici pdf nici imagine')
cover_url ='http://www.dilemaveche.ro/sites/all/themes/dilema/theme/dilema_two/layouter/dilema_two_homepage/logo.png'
return cover_url
def preprocess_html(self, soup):
return self.adeify_images(soup)