mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-07-09 03:04:10 -04:00
Introduce several news recipes
- nikkei
* free subscribe version works
* subscriber version need more work
- CNet/Japan
- The H
This commit is contained in:
Hiroshi Miura
parent
5f6ff5609d
commit
940d3fd6a6
30
resources/recipes/cnet.recipe
Normal file
30
resources/recipes/cnet.recipe
Normal file
@ -0,0 +1,30 @@
|
||||
import re;
|
||||
|
||||
class AdvancedUserRecipe1287588358(BasicNewsRecipe):
|
||||
title = u'CNET Japan'
|
||||
oldest_article = 3
|
||||
max_articles_per_feed = 30
|
||||
|
||||
feeds = [(u'cnet rss', u'http://feeds.japan.cnet.com/cnet/rss')]
|
||||
language = 'ja'
|
||||
encoding = 'Shift_JIS'
|
||||
remove_javascript = True
|
||||
|
||||
preprocess_regexps = [
|
||||
(re.compile(ur'<!--\u25B2contents_left END\u25B2-->.*</body>', re.DOTALL|re.IGNORECASE|re.UNICODE),
|
||||
lambda match: '</body>'),
|
||||
(re.compile(r'<!--AD_ELU_HEADER-->.*</body>', re.DOTALL|re.IGNORECASE),
|
||||
lambda match: '</body>'),
|
||||
(re.compile(ur'<!-- \u25B2\u95A2\u9023\u30BF\u30B0\u25B2 -->.*<!-- \u25B2ZDNet\u25B2 -->', re.UNICODE),
|
||||
lambda match: '<!-- removed -->'),
|
||||
]
|
||||
|
||||
remove_tags_before = dict(name="h2")
|
||||
remove_tags = [
|
||||
{'class':"social_bkm_share"},
|
||||
{'class':"social_bkm_print"},
|
||||
{'class':"block20 clearfix"},
|
||||
dict(name="div",attrs={'id':'bookreview'}),
|
||||
]
|
||||
remove_tags_after = {'class':"block20"}
|
||||
|
||||
48
resources/recipes/nikkei_free.recipe
Normal file
48
resources/recipes/nikkei_free.recipe
Normal file
@ -0,0 +1,48 @@
|
||||
class AdvancedUserRecipe1287958571(BasicNewsRecipe):
|
||||
title = u'\u65e5\u7d4c\u65b0\u805e\u96fb\u5b50\u7248(Free)'
|
||||
__author__ = 'Hiroshi Miura'
|
||||
description = 'News and current market affairs from Japan'
|
||||
oldest_article = 2
|
||||
max_articles_per_feed = 20
|
||||
language = 'ja'
|
||||
|
||||
feeds = [ (u'\u65e5\u7d4c\u4f01\u696d', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sangyo'),
|
||||
(u'\u65e5\u7d4c\u88fd\u54c1', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=newpro'),
|
||||
(u'internet', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=internet'),
|
||||
(u'\u653f\u6cbb', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=seiji'),
|
||||
(u'\u8ca1\u52d9', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=zaimu'),
|
||||
(u'\u7d4c\u6e08', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=keizai'),
|
||||
(u'\u56fd\u969b', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kaigai'),
|
||||
(u'\u79d1\u5b66', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kagaku'),
|
||||
(u'\u30de\u30fc\u30b1\u30c3\u30c8', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=market'),
|
||||
(u'\u304f\u3089\u3057', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kurashi'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sports'),
|
||||
(u'\u793e\u4f1a', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=shakai'),
|
||||
(u'\u30a8\u30b3', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=eco'),
|
||||
(u'\u5065\u5eb7', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kenkou'),
|
||||
(u'\u96c7\u7528', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=koyou'),
|
||||
(u'\u6559\u80b2', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kyouiku'),
|
||||
(u'\u304a\u304f\u3084\u307f', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=okuyami'),
|
||||
(u'\u4eba\u4e8b', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=zinzi'),
|
||||
(u'\u7279\u96c6', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=special'),
|
||||
(u'\u5730\u57df\u30cb\u30e5\u30fc\u30b9', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=local'),
|
||||
(u'\u7d71\u8a08\u30fb\u767d\u66f8', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=report'),
|
||||
(u'\u30e9\u30f3\u30ad\u30f3\u30b0', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=ranking'),
|
||||
(u'\u4f1a\u898b', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=interview'),
|
||||
(u'\u793e\u8aac\u30fb\u6625\u79cb', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=shasetsu'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30d7\u30ed\u91ce\u7403', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=baseball'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u5927\u30ea\u30fc\u30b0', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=mlb'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30b5\u30c3\u30ab\u30fc', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=soccer'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30b4\u30eb\u30d5', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=golf'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u76f8\u64b2', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sumou'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u7af6\u99ac', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=keiba'),
|
||||
(u'\u8abf\u67fb\u30fb\u30a2\u30f3\u30b1\u30fc\u30c8', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=research')
|
||||
]
|
||||
|
||||
remove_tags_before = dict(id="CONTENTS")
|
||||
remove_tags = [
|
||||
dict(name="form"),
|
||||
{'class':"cmn-hide"},
|
||||
]
|
||||
remove_tags_after = {'class':"cmn-pr_list"}
|
||||
|
||||
48
resources/recipes/nikkei_sports_sub.recipe
Normal file
48
resources/recipes/nikkei_sports_sub.recipe
Normal file
@ -0,0 +1,48 @@
|
||||
import string, re, sys
|
||||
from calibre import strftime
|
||||
from calibre.web.feeds.recipes import BasicNewsRecipe
|
||||
|
||||
class NikkeiNet_subscription(BasicNewsRecipe):
|
||||
title = u'\u65e5\u7d4c\u65b0\u805e\u96fb\u5b50\u7248'
|
||||
__author__ = 'Hiroshi Miura'
|
||||
description = 'News and current market affairs from Japan'
|
||||
needs_subscription = True
|
||||
oldest_article = 3
|
||||
max_articles_per_feed = 20
|
||||
language = 'ja'
|
||||
|
||||
feeds = [
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30d7\u30ed\u91ce\u7403', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=baseball'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u5927\u30ea\u30fc\u30b0', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=mlb'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30b5\u30c3\u30ab\u30fc', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=soccer'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30b4\u30eb\u30d5', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=golf'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u76f8\u64b2', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sumou'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u7af6\u99ac', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=keiba')
|
||||
]
|
||||
|
||||
remove_tags_before = dict(id="CONTENTS")
|
||||
remove_tags = [
|
||||
dict(name="form"),
|
||||
{'class':"cmn-hide"},
|
||||
]
|
||||
remove_tags_after = {'class':"cmn-pr_list"}
|
||||
recursions = 4
|
||||
|
||||
def get_browser(self):
|
||||
br = BasicNewsRecipe.get_browser()
|
||||
if self.username is not None and self.password is not None:
|
||||
br.open('https://id.nikkei.com/lounge/nl/base/LA0010.seam?cid=123456&flashId=654321')
|
||||
assert br.viewing_html()
|
||||
print br.title()
|
||||
response = br.response()
|
||||
response.set_data(response.get_data().replace("<input id=\"j_id48\"", "<!-- "))
|
||||
response.set_data(response.get_data().replace("gm_home_on.gif\" />", " -->"))
|
||||
br.set_response(response)
|
||||
br.select_form(name='LA0010Form01')
|
||||
br['LA0010Form01:LA0010Email'] = self.username
|
||||
br['LA0010Form01:LA0010Password'] = self.password
|
||||
res = br.submit()
|
||||
raw = res.read()
|
||||
if '日経IDのサービス一覧へ' not in raw:
|
||||
raise ValueError('Failed to log in to nikkei.net, check your username(email address) and password')
|
||||
return br
|
||||
71
resources/recipes/nikkei_sub.recipe
Normal file
71
resources/recipes/nikkei_sub.recipe
Normal file
@ -0,0 +1,71 @@
|
||||
import string, re, sys
|
||||
from calibre import strftime
|
||||
from calibre.web.feeds.recipes import BasicNewsRecipe
|
||||
|
||||
class NikkeiNet_subscription(BasicNewsRecipe):
|
||||
title = u'\u65e5\u7d4c\u65b0\u805e\u96fb\u5b50\u7248'
|
||||
__author__ = 'Hiroshi Miura'
|
||||
description = 'News and current market affairs from Japan'
|
||||
needs_subscription = True
|
||||
oldest_article = 2
|
||||
max_articles_per_feed = 20
|
||||
language = 'ja'
|
||||
recursions = 3
|
||||
remove_javascript = False
|
||||
|
||||
feeds = [ (u'\u65e5\u7d4c\u4f01\u696d', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sangyo'),
|
||||
(u'\u65e5\u7d4c\u88fd\u54c1', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=newpro'),
|
||||
(u'internet', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=internet'),
|
||||
(u'\u653f\u6cbb', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=seiji'),
|
||||
(u'\u8ca1\u52d9', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=zaimu'),
|
||||
(u'\u7d4c\u6e08', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=keizai'),
|
||||
(u'\u56fd\u969b', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kaigai'),
|
||||
(u'\u79d1\u5b66', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kagaku'),
|
||||
(u'\u30de\u30fc\u30b1\u30c3\u30c8', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=market'),
|
||||
(u'\u304f\u3089\u3057', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kurashi'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sports'),
|
||||
(u'\u793e\u4f1a', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=shakai'),
|
||||
(u'\u30a8\u30b3', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=eco'),
|
||||
(u'\u5065\u5eb7', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kenkou'),
|
||||
(u'\u96c7\u7528', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=koyou'),
|
||||
(u'\u6559\u80b2', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=kyouiku'),
|
||||
(u'\u304a\u304f\u3084\u307f', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=okuyami'),
|
||||
(u'\u4eba\u4e8b', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=zinzi'),
|
||||
(u'\u7279\u96c6', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=special'),
|
||||
(u'\u5730\u57df\u30cb\u30e5\u30fc\u30b9', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=local'),
|
||||
(u'\u7d71\u8a08\u30fb\u767d\u66f8', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=report'),
|
||||
(u'\u30e9\u30f3\u30ad\u30f3\u30b0', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=ranking'),
|
||||
(u'\u4f1a\u898b', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=interview'),
|
||||
(u'\u793e\u8aac\u30fb\u6625\u79cb', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=shasetsu'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30d7\u30ed\u91ce\u7403', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=baseball'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u5927\u30ea\u30fc\u30b0', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=mlb'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30b5\u30c3\u30ab\u30fc', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=soccer'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u30b4\u30eb\u30d5', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=golf'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u76f8\u64b2', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=sumou'),
|
||||
(u'\u30b9\u30dd\u30fc\u30c4\uff1a\u7af6\u99ac', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=keiba'),
|
||||
(u'\u8abf\u67fb\u30fb\u30a2\u30f3\u30b1\u30fc\u30c8', u'http://www.zou3.net/php/rss/nikkei2rss.php?head=research')
|
||||
]
|
||||
|
||||
def get_browser(self):
|
||||
br = BasicNewsRecipe.get_browser()
|
||||
if self.username is not None and self.password is not None:
|
||||
br.open('https://id.nikkei.com/lounge/nl/base/LA0010.seam')
|
||||
response = br.response()
|
||||
response.set_data(response.get_data().replace("<input id=\"j_id48\"", "<!-- "))
|
||||
response.set_data(response.get_data().replace("gm_home_on.gif\" />", " -->"))
|
||||
br.set_response(response)
|
||||
br.select_form(name='LA0010Form01')
|
||||
br['LA0010Form01:LA0010Email'] = self.username
|
||||
br['LA0010Form01:LA0010Password'] = self.password
|
||||
res = br.submit()
|
||||
raw = res.read()
|
||||
if '日経IDのサービス一覧へ' not in raw:
|
||||
raise ValueError('Failed to log in to nikkei.net, check your username(email address) and password')
|
||||
br.open('http://www.nikkei.com/')
|
||||
br.select_form(nr=0)
|
||||
res = br.submit()
|
||||
print res.read()
|
||||
return br
|
||||
|
||||
|
||||
|
||||
23
resources/recipes/the_h.recipe
Normal file
23
resources/recipes/the_h.recipe
Normal file
@ -0,0 +1,23 @@
|
||||
class AdvancedUserRecipe1289003166(BasicNewsRecipe):
|
||||
title = u'The H'
|
||||
__author__ = 'Hiroshi Miura'
|
||||
oldest_article = 3
|
||||
description = 'In association with Heise Online'
|
||||
publisher = 'Heise Media UK Ltd.'
|
||||
category = 'news, technology, security'
|
||||
max_articles_per_feed = 100
|
||||
language = 'en'
|
||||
encoding = 'utf-8'
|
||||
conversion_options = {
|
||||
'comment' : description
|
||||
,'tags' : category
|
||||
,'publisher': publisher
|
||||
,'language' : language
|
||||
}
|
||||
feeds = [
|
||||
(u'The H News Feed', u'http://www.h-online.com/news/atom.xml')
|
||||
]
|
||||
|
||||
def print_version(self, url):
|
||||
return url + '?view=print'
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user