[fixed recipe] Wprost - polish newsmagazine

[zaslav]

Below is fixed Wprost recipe - original recipe did not remove <table> tags - it causes problems on Kindle reader.

#!/usr/bin/env  python

__license__   = 'GPL v3'
__copyright__ = '2010, matek09, matek09@gmail.com'
__copyright__ = 'Modified 2011,  Mariusz Wolek <mariusz_dot_wolek @ gmail dot com>'

from calibre.web.feeds.news import BasicNewsRecipe
import re

class Wprost(BasicNewsRecipe):
        EDITION = 0
        FIND_LAST_FULL_ISSUE = True
        EXCLUDE_LOCKED = True
        ICO_BLOCKED = 'http://www.wprost.pl/G/icons/ico_blocked.gif'

        title = u'Wprost'
        __author__ = 'matek09'
        description = 'Weekly magazine'
        encoding = 'ISO-8859-2'
        no_stylesheets = True
        language = 'pl'
        remove_javascript = True

        remove_tags_before = dict(dict(name = 'div', attrs = {'id' : 'print-layer'}))
        remove_tags_after = dict(dict(name = 'div', attrs = {'id' : 'print-layer'}))

        '''keep_only_tags =[]
        keep_only_tags.append(dict(name = 'table', attrs = {'id' : 'title-table'}))
        keep_only_tags.append(dict(name = 'div', attrs = {'class' : 'div-header'}))
        keep_only_tags.append(dict(name = 'div', attrs = {'class' : 'div-content'}))
        keep_only_tags.append(dict(name = 'div', attrs = {'class' : 'def element-autor'}))'''

        preprocess_regexps = [(re.compile(r'style="display: none;"'), lambda match: ''),
        (re.compile(r'display: block;'), lambda match: ''),
        (re.compile(r'\<td\>\<tr\>\<\/table\>'), lambda match: ''),
        (re.compile(r'\<table .*?\>'), lambda match: ''),
        (re.compile(r'\<tr>'), lambda match: ''),
        (re.compile(r'\<td .*?\>'), lambda match: '')]

        remove_tags =[]
        remove_tags.append(dict(name = 'div', attrs = {'class' : 'def element-date'}))
        remove_tags.append(dict(name = 'div', attrs = {'class' : 'def silver'}))
        remove_tags.append(dict(name = 'div', attrs = {'id' : 'content-main-column-right'}))

        extra_css = '''
                                        .div-header {font-size: x-small; font-weight: bold}
                                        '''
#h2 {font-size: x-large; font-weight: bold}
        def is_blocked(self, a):
                if a.findNextSibling('img') is None:
                        return False
                else:
                        return True



        def find_last_issue(self):
                soup = self.index_to_soup('http://www.wprost.pl/archiwum/')
                a = 0
                if self.FIND_LAST_FULL_ISSUE:
                        ico_blocked = soup.findAll('img', attrs={'src' : self.ICO_BLOCKED})
                        a = ico_blocked[-1].findNext('a', attrs={'title' : re.compile('Zobacz spis tre.ci')})
                else:
                        a = soup.find('a', attrs={'title' : re.compile('Zobacz spis tre.ci')})
                self.EDITION = a['href'].replace('/tygodnik/?I=', '')
                self.cover_url = a.img['src']



        def parse_index(self):
                self.find_last_issue()
                soup = self.index_to_soup('http://www.wprost.pl/tygodnik/?I=' + self.EDITION)
                feeds = []
                for main_block in soup.findAll(attrs={'class':'main-block-s3 s3-head head-red3'}):
                        articles = list(self.find_articles(main_block))
                        if len(articles) > 0:
                                section = self.tag_to_string(main_block)
                                feeds.append((section, articles))
                return feeds

        def find_articles(self, main_block):
                for a in main_block.findAllNext( attrs={'style':['','padding-top: 15px;']}):
                        if a.name in "td":
                                break
                        if self.EXCLUDE_LOCKED & self.is_blocked(a):
                                continue
                        yield {
                                'title' : self.tag_to_string(a),
                                'url'   : 'http://www.wprost.pl' + a['href'],
                                'date'  : '',
                                'description' : ''
                                }