mirror of
https://github.com/kovidgoyal/calibre.git
synced 2025-06-23 15:30:45 -04:00
34 lines
1.2 KiB
Plaintext
34 lines
1.2 KiB
Plaintext
# -*- coding: utf8 -*-
|
|
|
|
from calibre.web.feeds.recipes import BasicNewsRecipe
|
|
import re
|
|
|
|
|
|
class Arcadia_BBS(BasicNewsRecipe):
|
|
title = u'Arcadia'
|
|
__author__ = 'Masahiro Hasegawa'
|
|
language = 'ja'
|
|
encoding = 'utf8'
|
|
filter_regexps = [r'ad\.jp\.ap\.valuecommerce.com', ]
|
|
timefmt = '[%Y/%m/%d]'
|
|
remove_tags_before = dict(name='a', attrs={'name': 'kiji'})
|
|
|
|
sid_list = [] # some sotory id
|
|
|
|
def parse_index(self):
|
|
result = []
|
|
for sid in self.sid_list:
|
|
s_result = []
|
|
soup = self.index_to_soup(
|
|
'http://www.mai-net.net/bbs/sst/sst.php?act=dump&all=%d'
|
|
% sid)
|
|
sec = soup.findAll('a', attrs={'href': re.compile(r'.*?kiji')})
|
|
for s in sec[:-2]:
|
|
s_result.append(dict(title=s.string,
|
|
url="http://www.mai-net.net" + s['href'],
|
|
date=s.parent.parent.parent.findAll('td')[
|
|
3].string[:-6],
|
|
description='', content=''))
|
|
result.append((s_result[0]['title'], s_result))
|
|
return result
|