Arcadia by Masahiro Hasegawa

This commit is contained in:
Kovid Goyal 2012-08-28 01:28:49 +05:30
parent 955990c17d
commit 6c5d581218

35
recipes/arcadia.recipe Normal file
View File

@ -0,0 +1,35 @@
# -*- 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