mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-07-09 03:04:10 -04:00
New recipes for HRT and RTS by Darko Miletic
This commit is contained in:
Kovid Goyal
parent
1a1cf7f1b9
commit
1c56b03f95
BIN
src/calibre/gui2/images/news/hrt.png
Normal file
BIN
src/calibre/gui2/images/news/hrt.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 606 B |
BIN
src/calibre/gui2/images/news/rts.png
Normal file
BIN
src/calibre/gui2/images/news/rts.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 458 B |
@ -42,7 +42,7 @@ recipe_modules = ['recipe_' + r for r in (
|
|||||||
'moneynews', 'der_standard', 'diepresse', 'nzz_ger', 'hna',
|
'moneynews', 'der_standard', 'diepresse', 'nzz_ger', 'hna',
|
||||||
'seattle_times', 'scott_hanselman', 'coding_horror', 'twitchfilms',
|
'seattle_times', 'scott_hanselman', 'coding_horror', 'twitchfilms',
|
||||||
'stackoverflow', 'telepolis_artikel', 'zaobao', 'usnews',
|
'stackoverflow', 'telepolis_artikel', 'zaobao', 'usnews',
|
||||||
'straitstimes', 'index_hu', 'pcworld_hu',
|
'straitstimes', 'index_hu', 'pcworld_hu', 'hrt', 'rts',
|
||||||
)]
|
)]
|
||||||
|
|
||||||
import re, imp, inspect, time, os
|
import re, imp, inspect, time, os
|
||||||
|
|||||||
66
src/calibre/web/feeds/recipes/recipe_hrt.py
Normal file
66
src/calibre/web/feeds/recipes/recipe_hrt.py
Normal file
@ -0,0 +1,66 @@
|
|||||||
|
#!/usr/bin/env python
|
||||||
|
|
||||||
|
__license__ = 'GPL v3'
|
||||||
|
__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
|
||||||
|
|
||||||
|
'''
|
||||||
|
www.hrt.hr
|
||||||
|
'''
|
||||||
|
|
||||||
|
import re
|
||||||
|
from calibre.web.feeds.news import BasicNewsRecipe
|
||||||
|
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag
|
||||||
|
|
||||||
|
class HRT(BasicNewsRecipe):
|
||||||
|
title = 'HRT: Vesti'
|
||||||
|
__author__ = 'Darko Miletic'
|
||||||
|
description = 'News from Croatia'
|
||||||
|
publisher = 'HRT'
|
||||||
|
category = 'news, politics, Croatia, HRT'
|
||||||
|
no_stylesheets = True
|
||||||
|
encoding = 'utf-8'
|
||||||
|
use_embedded_content = False
|
||||||
|
language = _("Croatian")
|
||||||
|
lang = 'hr-HR'
|
||||||
|
extra_css = '@font-face {font-family: "serif1";src:url(res:///opt/sony/ebook/FONT/tt0011m_.ttf)} body{font-family: serif1, serif} .article_description{font-family: serif1, serif}'
|
||||||
|
|
||||||
|
html2lrf_options = [
|
||||||
|
'--comment', description
|
||||||
|
, '--category', category
|
||||||
|
, '--publisher', publisher
|
||||||
|
]
|
||||||
|
|
||||||
|
html2epub_options = 'publisher="' + publisher + '"\ncomments="' + description + '"\ntags="' + category + '"\noverride_css=" p {text-indent: 0em; margin-top: 0em; margin-bottom: 0.5em} img {margin-top: 0em; margin-bottom: 0.4em}"'
|
||||||
|
|
||||||
|
|
||||||
|
preprocess_regexps = [(re.compile(u'\u0110'), lambda match: u'\u00D0')]
|
||||||
|
|
||||||
|
keep_only_tags = [dict(name='div', attrs={'class':'bigVijest'})]
|
||||||
|
|
||||||
|
remove_tags = [dict(name=['object','link','embed'])]
|
||||||
|
|
||||||
|
remove_tags_after = dict(name='div', attrs={'class':'nsAuthor'})
|
||||||
|
|
||||||
|
feeds = [
|
||||||
|
(u'Vijesti' , u'http://www.hrt.hr/?id=316&type=100&rss=vijesti' )
|
||||||
|
,(u'Sport' , u'http://www.hrt.hr/?id=316&type=100&rss=sport' )
|
||||||
|
,(u'Zabava' , u'http://www.hrt.hr/?id=316&type=100&rss=zabava' )
|
||||||
|
,(u'Filmovi i serije' , u'http://www.hrt.hr/?id=316&type=100&rss=filmovi' )
|
||||||
|
,(u'Dokumentarni program', u'http://www.hrt.hr/?id=316&type=100&rss=dokumentarci')
|
||||||
|
,(u'Glazba' , u'http://www.hrt.hr/?id=316&type=100&rss=glazba' )
|
||||||
|
,(u'Kultura' , u'http://www.hrt.hr/?id=316&type=100&rss=kultura' )
|
||||||
|
,(u'Mladi' , u'http://www.hrt.hr/?id=316&type=100&rss=mladi' )
|
||||||
|
,(u'Manjine' , u'http://www.hrt.hr/?id=316&type=100&rss=manjine' )
|
||||||
|
,(u'Radio' , u'http://www.hrt.hr/?id=316&type=100&rss=radio' )
|
||||||
|
]
|
||||||
|
|
||||||
|
def preprocess_html(self, soup):
|
||||||
|
soup.html['xml:lang'] = self.lang
|
||||||
|
soup.html['lang'] = self.lang
|
||||||
|
mlang = Tag(soup,'meta',[("http-equiv","Content-Language"),("content",self.lang)])
|
||||||
|
mcharset = Tag(soup,'meta',[("http-equiv","Content-Type"),("content","text/html; charset=UTF-8")])
|
||||||
|
soup.head.insert(0,mlang)
|
||||||
|
soup.head.insert(1,mcharset)
|
||||||
|
for item in soup.findAll(style=True):
|
||||||
|
del item['style']
|
||||||
|
return self.adeify_images(soup)
|
||||||
60
src/calibre/web/feeds/recipes/recipe_rts.py
Normal file
60
src/calibre/web/feeds/recipes/recipe_rts.py
Normal file
@ -0,0 +1,60 @@
|
|||||||
|
#!/usr/bin/env python
|
||||||
|
|
||||||
|
__license__ = 'GPL v3'
|
||||||
|
__copyright__ = '2009, Darko Miletic <darko.miletic at gmail.com>'
|
||||||
|
|
||||||
|
'''
|
||||||
|
www.rts.rs
|
||||||
|
'''
|
||||||
|
|
||||||
|
import re
|
||||||
|
from calibre.web.feeds.news import BasicNewsRecipe
|
||||||
|
from calibre.ebooks.BeautifulSoup import BeautifulSoup, Tag
|
||||||
|
|
||||||
|
class RTS(BasicNewsRecipe):
|
||||||
|
title = 'RTS: Vesti'
|
||||||
|
__author__ = 'Darko Miletic'
|
||||||
|
description = 'News from Serbia'
|
||||||
|
publisher = 'RTS'
|
||||||
|
category = 'news, politics, Serbia, RTS'
|
||||||
|
no_stylesheets = True
|
||||||
|
encoding = 'utf-8'
|
||||||
|
use_embedded_content = True
|
||||||
|
language = _("Serbian")
|
||||||
|
lang = 'sr-Latn-RS'
|
||||||
|
extra_css = '@font-face {font-family: "serif1";src:url(res:///opt/sony/ebook/FONT/tt0011m_.ttf)} body{font-family: serif1, serif} .article_description{font-family: serif1, serif}'
|
||||||
|
|
||||||
|
html2lrf_options = [
|
||||||
|
'--comment', description
|
||||||
|
, '--category', category
|
||||||
|
, '--publisher', publisher
|
||||||
|
]
|
||||||
|
|
||||||
|
html2epub_options = 'publisher="' + publisher + '"\ncomments="' + description + '"\ntags="' + category + '"\noverride_css=" p {text-indent: 0em; margin-top: 0em; margin-bottom: 0.5em} img {margin-top: 0em; margin-bottom: 0.4em}"'
|
||||||
|
|
||||||
|
|
||||||
|
preprocess_regexps = [(re.compile(u'\u0110'), lambda match: u'\u00D0')]
|
||||||
|
|
||||||
|
feeds = [
|
||||||
|
(u'Vesti' , u'http://www.rts.rs/page/stories/sr/rss.html' )
|
||||||
|
,(u'Srbija' , u'http://www.rts.rs/page/stories/sr/rss/9/Srbija.html' )
|
||||||
|
,(u'Region' , u'http://www.rts.rs/page/stories/sr/rss/11/Region.html' )
|
||||||
|
,(u'Svet' , u'http://www.rts.rs/page/stories/sr/rss/10/Svet.html' )
|
||||||
|
,(u'Hronika' , u'http://www.rts.rs/page/stories/sr/rss/135/Hronika.html' )
|
||||||
|
,(u'Drustvo' , u'http://www.rts.rs/page/stories/sr/rss/125/Dru%C5%A1tvo.html')
|
||||||
|
,(u'Ekonomija' , u'http://www.rts.rs/page/stories/sr/rss/13/Ekonomija.html' )
|
||||||
|
,(u'Nauka' , u'http://www.rts.rs/page/stories/sr/rss/14/Nauka.html' )
|
||||||
|
,(u'Kultura' , u'http://www.rts.rs/page/stories/sr/rss/16/Kultura.html' )
|
||||||
|
,(u'Zanimljivosti' , u'http://www.rts.rs/page/stories/sr/rss/15/Zanimljivosti.html')
|
||||||
|
,(u'Sport' , u'http://www.rts.rs/page/sport/sr/rss.html' )
|
||||||
|
]
|
||||||
|
|
||||||
|
def preprocess_html(self, soup):
|
||||||
|
soup.html['xml:lang'] = self.lang
|
||||||
|
soup.html['lang'] = self.lang
|
||||||
|
mlang = Tag(soup,'meta',[("http-equiv","Content-Language"),("content",self.lang)])
|
||||||
|
mcharset = Tag(soup,'meta',[("http-equiv","Content-Type"),("content","text/html; charset=UTF-8")])
|
||||||
|
soup.head.insert(0,mlang)
|
||||||
|
soup.head.insert(1,mcharset)
|
||||||
|
return self.adeify_images(soup)
|
||||||
|
|
||||||
Loading…
x
Reference in New Issue
Block a user