calibre/recipes/mediapart.recipe

#!/usr/bin/env python2
# vim:fileencoding=utf-8
from __future__ import unicode_literals

__license__ = 'GPL v3'
__copyright__ = '2016, Daniel Bonnery ? (contact: DanielBonnery sur mobileread.com) 2009, Mathieu Godlewski <mathieu at godlewski.fr>; 2010-2012, Louis Gesbert <meta at antislash dot info>'  # noqa
'''
Mediapart
'''

import re
from calibre.web.feeds.news import BasicNewsRecipe
from calibre.web.feeds import feeds_from_index
from datetime import date, timedelta


class Mediapart(BasicNewsRecipe):
    title = 'Mediapart'
    __author__ = 'Daniel Bonnery from a version by Mathieu Godlewski, Louis Gesbert'
    description = 'Global news in french from news site Mediapart'
    publication_type = 'newspaper'
    language = 'fr'
    needs_subscription = True
    oldest_article = 2

    use_embedded_content = False
    no_stylesheets = True

    cover_url = 'https://static.mediapart.fr/files/M%20Philips/logo-mediapart.png'

# --

    oldest_article_date = date.today() - timedelta(days=oldest_article)

# -- get the index (the feed at 'http://www.mediapart.fr/articles/feed' only has
#    the 10 last elements :/)

    feeds = [
        ('La Une', 'http://www.mediapart.fr/articles/feed'),
    ]

    def parse_feeds(self):
        feeds = super(Mediapart, self).parse_feeds()
        feeds += feeds_from_index(self.my_parse_index(feeds))
        return feeds

    def my_parse_index(self, la_une):
        articles = []

        breves = []
        liens = []
        confidentiels = []

        soup = self.index_to_soup(
            'https://www.mediapart.fr/journal/fil-dactualites')
        page = soup.find('main', {'class': 'global-wrapper'})
        fils = page.find('ul', {'class': 'post-list universe-journal'})

        for article in fils.findAll('li'):
            try:
                title = article.find('h3', recursive=False)

                if title is None or title['class'] == 'title-specific':
                    continue

                # print "found fil ",title
                article_type = article.find('a', {'href': re.compile(
                    r'.*\/type-darticles\/.*')}).renderContents()
                # print "kind: ",article_type

                for s in title('span'):
                    s.replaceWith(s.renderContents() + "\n")
                url = title.find('a', href=True)['href']

                # article_date = self.parse_french_date(article.find("span", "article-date").renderContents())
                # print("################################# 9")
                # print(article_date)

                # if article_date < self.oldest_article_date:
                #    print "too old"
                #    continue

                authors = article.findAll(
                    'a', {'class': re.compile(r'\bjournalist\b')})
                authors = [self.tag_to_string(a) for a in authors]

                # description = article.find('div', {'class': lambda c: c != 'taxonomy-teaser'}, recursive=False).findAll('p')

                # print "fil ",title," by ",authors," : ",description

                summary = {
                    'title': self.tag_to_string(title).strip(),
                    'author': ', '.join(authors),
                    'url': 'https://www.mediapart.fr' + url
                }
                if article_type == 'Lien':
                    liens.append(summary)
                if article_type == 'Confidentiel':
                    confidentiels.append(summary)
                if article_type not in ['Lien', 'Confidentiel']:
                    breves.append(summary)
            except:
                pass

        # print 'La Une: ', len(la_une), ' articles'
        # for a in la_une: print a["title"]
        # print 'Brèves: ', len(breves), ' articles'
        # print 'Revue web: ', len(liens), ' articles'
        # print 'Confidentiel: ', len(confidentiels), ' articles'

        articles += [('Brèves', breves)] if breves else []
        articles += [('Revue du Web', liens)] if liens else []
        articles += [('Confidentiel', confidentiels)] if confidentiels else []
        return articles
# -- print-version

    conversion_options = {'smarten_punctuation': True}

    remove_tags = [dict(name='div', attrs={'class': 'print-source_url'})]

    # non-locale specific date parse (strptime("%d %b %Y",s) would work with
    # french locale)
    def parse_french_date(self, date_str):
        date_arr = date_str.lower().split()
        return date(day=int(date_arr[0]),
                    year=int(date_arr[2]),
                    month=[None, 'janvier', 'février', 'mars', 'avril', 'mai', 'juin', 'juillet',
                           'août', 'septembre', 'octobre', 'novembre', 'décembre'].index(date_arr[1]))

    def print_version(self, url):
        soup = self.index_to_soup(url)
        # Filter old articles
        # article_date = self.parse_french_date(self.tag_to_string(soup.find('span', 'article-date')))

        # if article_date < self.oldest_article_date:
        #     return None

        tools = soup.find('li', {'class': 'print'})
        link = tools.find('a', {'href': re.compile(r'\/print\/.*')})
        # if link is None:
        #    print 'Error: print link not found'
        #    return None
        return 'https://mediapart.fr' + link['href']

# -- Handle login
    def get_browser(self):
        def is_form_login(form):
            return "id" in form.attrs and form.attrs['id'] == "logFormEl"
        br = BasicNewsRecipe.get_browser(self)
        if self.username is not None and self.password is not None:
            br.open('https://www.mediapart.fr/login')
            br.select_form(predicate=is_form_login)
            br['name'] = self.username
            br['password'] = self.password
            br.submit()
        return br

    # This is a workaround articles with scribd content that include
    # <body></body> tags _within_ the body
    preprocess_regexps = [
        (re.compile(r'(<body.*?>)(.*)</body>', re.IGNORECASE | re.DOTALL),
         lambda match:
             match.group(1) + re.sub(
                 re.compile(r'</?body>', re.IGNORECASE | re.DOTALL), '', match.group(2)) + '</body>')
    ]