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

Merge branch 'granta' of https://github.com/Dhar/calibre

Browse Source
This commit is contained in:
Kovid Goyal 2018-03-06 06:02:37 +05:30
commit 3451698bf3
No known key found for this signature in database
GPG Key ID: 06BC317B515ACE7C
1 changed files with 12 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
recipes/granta.recipe
View File

@ -159,17 +159,21 @@ def absurl(url):
url = 'https://www.granta.com' + url url = 'https://www.granta.com' + url
return url return url
def stripstyle(tag): def stripstyle(tag):
if tag is not None: if tag is not None:
del tag['style'] del tag['style']
def get_innermost_string(tag):
while hasattr(tag, 'contents') and len(tag.contents) > 0 and tag.contents[0] is not None:
tag = tag.contents[0]
return str(tag).strip()
################################################################## ##################################################################
class Granta(BasicNewsRecipe): class Granta(BasicNewsRecipe):
title = u'Granta' title = u'Granta'
description = u'Granta magazine' description = u'The Magazine of New Writing'
language = 'en' language = 'en'
__author__ = 'Gary Arnold' __author__ = 'Gary Arnold'
@ -266,17 +270,17 @@ class Granta(BasicNewsRecipe):
# Either user is logged in or the article is unlocked # Either user is logged in or the article is unlocked
h1 = toc.find('h1') h1 = toc.find('h1')
h2 = toc.find('h2') h2 = toc.find('h2')
if h1.find('a') is not None and h1.find('a').contents is not None: if h1.find('a') is not None and len(h1.find('a').contents) > 0 and h1.find('a').contents[0] is not None:
title = h1.find('a').contents[0].strip() title = get_innermost_string(h1.find('a').contents[0])
elif len(h1.contents) > 0 and h1.contents[0] is not None: elif len(h1.contents) > 0 and h1.contents[0] is not None:
title = h1.contents[0] title = get_innermost_string(h1.contents[0])
else: else:
title = '' title = ''
if h2.find('a') is not None and h2.find('a').contents is not None: if h2.find('a') is not None and len(h2.find('a').contents) > 0 and h2.find('a').contents[0] is not None:
author = h2.find('a').contents[0].strip() author = get_innermost_string(h2.find('a').contents[0])
title = title + u' (%s)' % author title = title + u' (%s)' % author
elif len(h2.contents) > 0 and h2.contents[0] is not None: elif len(h2.contents) > 0 and h2.contents[0] is not None:
author = h2.contents[0] author = get_innermost_string(h2.contents[0])
title = title + u' (%s)' % author title = title + u' (%s)' % author
else: else:
author = '' author = ''