Economist update

This commit is contained in:
unkn0w7n 2023-12-08 13:18:15 +05:30
parent d20b1a08db
commit 734c52ee87
4 changed files with 15 additions and 75 deletions

View File

@ -200,31 +200,9 @@ class Economist(BasicNewsRecipe):
self.web2disk_options.compress_news_images_auto_size = 5 self.web2disk_options.compress_news_images_auto_size = 5
self.log.warn('Kindle Output profile being used, reducing image quality to keep file size below amazon email threshold') self.log.warn('Kindle Output profile being used, reducing image quality to keep file size below amazon email threshold')
def get_browser(self): def get_browser(self, *args, **kwargs):
br = BasicNewsRecipe.get_browser(self) kwargs['user_agent'] = 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
# Add a cookie indicating we have accepted Economist's cookie br = BasicNewsRecipe.get_browser(self, *args, **kwargs)
# policy (needed when running from some European countries)
ck = Cookie(
version=0,
name='notice_preferences',
value='2:',
port=None,
port_specified=False,
domain='.economist.com',
domain_specified=False,
domain_initial_dot=True,
path='/',
path_specified=False,
secure=False,
expires=None,
discard=False,
comment=None,
comment_url=None,
rest={'HttpOnly': None},
rfc2109=False
)
br.cookiejar.set_cookie(ck)
br.set_handle_gzip(True)
return br return br
def preprocess_raw_html(self, raw, url): def preprocess_raw_html(self, raw, url):

View File

@ -63,3 +63,8 @@ class Espresso(BasicNewsRecipe):
nt = new_tag(soup, 'hr') nt = new_tag(soup, 'hr')
hr.append(nt) hr.append(nt)
return soup return soup
def get_browser(self, *args, **kwargs):
kwargs['user_agent'] = 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
br = BasicNewsRecipe.get_browser(self, *args, **kwargs)
return br

View File

@ -200,31 +200,9 @@ class Economist(BasicNewsRecipe):
self.web2disk_options.compress_news_images_auto_size = 5 self.web2disk_options.compress_news_images_auto_size = 5
self.log.warn('Kindle Output profile being used, reducing image quality to keep file size below amazon email threshold') self.log.warn('Kindle Output profile being used, reducing image quality to keep file size below amazon email threshold')
def get_browser(self): def get_browser(self, *args, **kwargs):
br = BasicNewsRecipe.get_browser(self) kwargs['user_agent'] = 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
# Add a cookie indicating we have accepted Economist's cookie br = BasicNewsRecipe.get_browser(self, *args, **kwargs)
# policy (needed when running from some European countries)
ck = Cookie(
version=0,
name='notice_preferences',
value='2:',
port=None,
port_specified=False,
domain='.economist.com',
domain_specified=False,
domain_initial_dot=True,
path='/',
path_specified=False,
secure=False,
expires=None,
discard=False,
comment=None,
comment_url=None,
rest={'HttpOnly': None},
rfc2109=False
)
br.cookiejar.set_cookie(ck)
br.set_handle_gzip(True)
return br return br
def preprocess_raw_html(self, raw, url): def preprocess_raw_html(self, raw, url):

View File

@ -196,31 +196,9 @@ class Economist(BasicNewsRecipe):
self.log.warn('Kindle Output profile being used, reducing image quality to keep file size below amazon email threshold') self.log.warn('Kindle Output profile being used, reducing image quality to keep file size below amazon email threshold')
def get_browser(self): def get_browser(self, *args, **kwargs):
br = BasicNewsRecipe.get_browser(self) kwargs['user_agent'] = 'facebookexternalhit/1.1 (+http://www.facebook.com/externalhit_uatext.php)'
# Add a cookie indicating we have accepted Economist's cookie br = BasicNewsRecipe.get_browser(self, *args, **kwargs)
# policy (needed when running from some European countries)
ck = Cookie(
version=0,
name='notice_preferences',
value='2:',
port=None,
port_specified=False,
domain='.economist.com',
domain_specified=False,
domain_initial_dot=True,
path='/',
path_specified=False,
secure=False,
expires=None,
discard=False,
comment=None,
comment_url=None,
rest={'HttpOnly': None},
rfc2109=False
)
br.cookiejar.set_cookie(ck)
br.set_handle_gzip(True)
return br return br
def preprocess_raw_html(self, raw, url): def preprocess_raw_html(self, raw, url):
@ -253,6 +231,7 @@ class Economist(BasicNewsRecipe):
x.set('style', 'text-align:center; font-size:small;') x.set('style', 'text-align:center; font-size:small;')
for x in root.xpath('//cite'): for x in root.xpath('//cite'):
x.tag = 'blockquote' x.tag = 'blockquote'
x.set('style', 'color:#404040;')
raw = etree.tostring(root, encoding='unicode') raw = etree.tostring(root, encoding='unicode')
return raw return raw