mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-09-29 15:31:08 -04:00
77 lines
2.1 KiB
Python
77 lines
2.1 KiB
Python
#!/usr/bin/env python2
|
|
# -*- coding: utf-8 -*-
|
|
#Based on veezh's original recipe and Kovid Goyal's New York Times recipe
|
|
|
|
__license__ = 'GPL v3'
|
|
__copyright__ = '2011, Snaab'
|
|
|
|
'''
|
|
www.nrc.nl
|
|
'''
|
|
import os, zipfile
|
|
import time
|
|
from calibre.web.feeds.news import BasicNewsRecipe
|
|
from calibre.ptempfile import PersistentTemporaryFile
|
|
|
|
|
|
class NRCHandelsblad(BasicNewsRecipe):
|
|
|
|
title = u'NRC Handelsblad'
|
|
description = u'De ePaper-versie van NRC'
|
|
language = 'nl'
|
|
lang = 'nl-NL'
|
|
needs_subscription = True
|
|
|
|
__author__ = 'Snaab'
|
|
|
|
conversion_options = {
|
|
'no_default_epub_cover' : True
|
|
}
|
|
|
|
def get_browser(self):
|
|
br = BasicNewsRecipe.get_browser()
|
|
if self.username is not None and self.password is not None:
|
|
br.open('http://login.nrc.nl/login')
|
|
br.select_form(nr=0)
|
|
br['username'] = self.username
|
|
br['password'] = self.password
|
|
br.submit()
|
|
return br
|
|
|
|
def build_index(self):
|
|
|
|
today = time.strftime("%Y%m%d")
|
|
|
|
domain = "http://digitaleeditie.nrc.nl"
|
|
|
|
url = domain + "/digitaleeditie/helekrant/epub/nrc_" + today + ".epub"
|
|
#print url
|
|
|
|
try:
|
|
br = self.get_browser()
|
|
f = br.open(url)
|
|
except:
|
|
self.report_progress(0,_('Kan niet inloggen om editie te downloaden'))
|
|
raise ValueError('Krant van vandaag nog niet beschikbaar')
|
|
|
|
|
|
tmp = PersistentTemporaryFile(suffix='.epub')
|
|
self.report_progress(0,_('downloading epub'))
|
|
tmp.write(f.read())
|
|
f.close()
|
|
br.close()
|
|
if zipfile.is_zipfile(tmp):
|
|
try:
|
|
zfile = zipfile.ZipFile(tmp.name, 'r')
|
|
zfile.extractall(self.output_dir)
|
|
self.report_progress(0,_('extracting epub'))
|
|
except zipfile.BadZipfile:
|
|
self.report_progress(0,_('BadZip error, continuing'))
|
|
|
|
tmp.close()
|
|
index = os.path.join(self.output_dir, 'metadata.opf')
|
|
|
|
self.report_progress(1,_('epub downloaded and extracted'))
|
|
|
|
return index
|