calibre/recipes/caravan_magazine.recipe
Kovid Goyal 567040ee1e Perform PEP8 compliance checks on the entire codebase
Some bits of PEP 8 are turned off via setup.cfg
2016-07-29 21:25:17 +05:30

143 lines
5.5 KiB
Plaintext

# coding: utf-8
import html5lib
import re
from lxml import etree
from calibre.web.feeds.recipes import BasicNewsRecipe
from calibre.utils.cleantext import clean_xml_chars
from calibre.ebooks.BeautifulSoup import Tag, NavigableString
class CaravanMagazine(BasicNewsRecipe):
title = 'Caravan Magazine'
__author__ = 'Kovid Goyal, Gobelinus'
description = 'An Indian Journal of politics and culture'
language = 'en_IN'
timefmt = ' [%b, %Y]'
no_stylesheets = True
keep_only_tags = [
dict(attrs={'class': ['post-title']}),
dict(attrs={'class': ['post-subhheading',
'authorndate', 'rg-thumbs', 'entry-content']}),
]
remove_tags = [
dict(attrs={'class': ['share-with']}),
]
def preprocess_raw_html(self, raw_html, url):
root = html5lib.parse(
clean_xml_chars(raw_html), treebuilder='lxml',
namespaceHTMLElements=False)
for s in root.xpath('//script'):
s.getparent().remove(s)
return etree.tostring(root, encoding=unicode)
def preprocess_html(self, soup):
# Handle the image carousel
carousel = soup.find('div', {'class': 'rg-thumbs'})
if carousel is not None:
# create a new container to collect all images
all_images = Tag(soup, 'div')
# all_images['class'] = 'rg-thumbs'
for index, img in enumerate(carousel.findAll('img')):
# create a new div to contain image and caption
div = Tag(soup, 'div')
div['style'] = 'text-align:left;font-size:70%;margin-bottom: 0.4em;'
ns = NavigableString(img['data-caption'])
img['src'] = img['data-large']
del img['data-large']
del img['data-caption']
del img['data-credit']
img.extract()
div.insert(0, img)
div.insert(1, Tag(soup, 'br'))
div.insert(3, ns)
div.insert(3, Tag(soup, 'br'))
all_images.insert(index, div)
# extracted all images, replace carousel with extracted images
carousel.replaceWith(all_images)
return soup
# To parse artice toc
def parse_index(self):
base_url = 'http://www.caravanmagazine.in'
raw = self.index_to_soup('{0}/current-issue'.format(base_url),
raw=True)
raw = raw.decode('utf-8')
raw = self.preprocess_raw_html(raw, None)
soup = self.index_to_soup(raw)
# find current issue cover
try:
cover_img = soup.find('div', {'class': 'issue-image'}).find('img')
# a = soup.find('a', rel=lambda x:x and '[field_c_issues_image]' in x)
# if a is not None:
self.cover_url = cover_img['src']
except:
pass
# ci = soup.find(attrs={'class': 'current-issue-block'})
ci = soup.findAll(attrs={'class': re.compile('archive-story.*')})
current_section = 'Section'
current_articles = []
feeds = []
# define some reusable constants
heading_class = 'subject-heading'
content_class = 'subject-content'
stories_re = re.compile('({0}|{1}).*'.format(heading_class,
content_class))
for story in ci:
for ele in story.findAll(attrs={'class': stories_re}):
if ele['class'].startswith(heading_class):
# heading section
if current_articles:
self.log('Adding {0} articles to {1}'.format(
len(current_articles), current_section))
feeds.append((current_section, current_articles))
current_section = self.tag_to_string(ele)
current_articles = []
self.log('Section:', current_section)
pass
else:
# content Section
for art in ele.findAll('article',
attrs={'id': re.compile('post-.*')}):
title = art.find('h1')
if title is not None:
a = title.find('a', href=True)
if a is not None:
href = a['href']
# convert relative url to absolute url
if href.startswith('/'):
href = '{0}{1}'.format(base_url, href)
article = {
'title': self.tag_to_string(title),
'url': href
}
title.extract()
desc = self.tag_to_string(art).strip()
if desc:
article['description'] = desc
current_articles.append(article)
self.log('\t' + article['title'])
self.log('\t\t' + article['url'])
# append any remaining articles that were probably from last section,
# we ran out of heading_class to push them
if current_articles:
self.log('Adding {0} articles to {1}'.format(
len(current_articles), current_section))
feeds.append((current_section, current_articles))
return feeds