对于Calibre,我正在努力修复built-in recipe for Vedomosti。
#!/usr/bin/env python
u'''
Ведомости
'''
from calibre.web.feeds.feedparser import parse
from calibre.ebooks.BeautifulSoup import Tag
from calibre.web.feeds.news import BasicNewsRecipe
class VedomostiRecipe(BasicNewsRecipe):
title = u'Ведомости'
__author__ = 'Nikolai Kotchetkov'
publisher = 'vedomosti.ru'
category = 'press, Russia'
description = u'Ежедневная деловая газета'
oldest_article = 3
max_articles_per_feed = 100
masthead_url = u'http://motorro.com/imgdir/logos/ved_logo_black2_cropped.gif'
cover_url = u'http://motorro.com/imgdir/logos/ved_logo_black2_cropped.gif'
#Add feed names if you want them to be sorted (feeds of this list appear first)
sortOrder = [u'_default', u'Первая полоса', u'Власть и деньги']
encoding = 'cp1251'
language = 'ru'
no_stylesheets = True
remove_javascript = True
recursions = 0
conversion_options = {
'comment' : description
, 'tags' : category
, 'publisher' : publisher
, 'language' : language
}
#keep_only_tags = [dict(name='td', attrs={'class' : ['second_content']})]
#remove_tags_after = [dict(name='div', attrs={'class' : 'article_text'})]
#remove_tags = [dict(name='div', attrs={'class' : ['sep', 'choice', 'articleRightTbl']})]
feeds = [u'http://www.vedomosti.ru/newspaper/out/rss.xml']
#base URL for relative links
base_url = u'http://www.vedomosti.ru'
extra_css = 'h1 {font-size: 1.5em; margin: 0em 0em 0em 0em; text-align: center;}'\
'h2 {font-size: 1.0em; margin: 0em 0em 0em 0em;}'\
'h3 {font-size: 0.8em; margin: 0em 0em 0em 0em;}'\
'.article_date {font-size: 0.5em; color: gray; font-family: monospace; text-align:right;}'\
'.article_authors {font-size: 0.5em; color: gray; font-family: monospace; text-align:right;}'\
'.article_img {width:100%; text-align: center; padding: 3px 3px 3px 3px;}'\
'.article_img_desc {width:100%; text-align: center; font-size: 0.5em; color: gray; font-family: monospace;}'\
'.article_desc {font-size: 1em; font-style:italic;}'
def parse_index(self):
try:
feedData = parse(self.feeds[0])
if not feedData:
raise NotImplementedError
self.log("parse_index: Feed loaded successfully.")
if feedData.feed.has_key('title'):
self.title = feedData.feed.title
self.log("parse_index: Title updated to: ", self.title)
if feedData.feed.has_key('description'):
self.description = feedData.feed.description
self.log("parse_index: Description updated to: ", self.description)
def get_virtual_feed_articles(feed):
if feeds.has_key(feed):
return feeds[feed][1]
self.log("Adding new feed: ", feed)
articles = []
feeds[feed] = (feed, articles)
return articles
feeds = {}
#Iterate feed items and distribute articles using tags
for item in feedData.entries:
link = item.get('link', '');
title = item.get('title', '');
if '' == link or '' == title:
continue
article = {'title':title, 'url':link, 'description':item.get('description', ''), 'date':item.get('date', ''), 'content':''};
if not item.has_key('tags'):
get_virtual_feed_articles('_default').append(article)
continue
for tag in item.tags:
addedToDefault = False
term = tag.get('term', '')
if '' == term:
if (not addedToDefault):
get_virtual_feed_articles('_default').append(article)
continue
get_virtual_feed_articles(term).append(article)
#Get feed list
#Select sorted feeds first of all
result = []
for feedName in self.sortOrder:
if (not feeds.has_key(feedName)): continue
result.append(feeds[feedName])
del feeds[feedName]
result = result + feeds.values()
return result
except Exception, err:
self.log(err)
raise NotImplementedError
def preprocess_html(self, soup):
return self.adeify_images(soup)
def postprocess_html(self, soup, first_fetch):
#self.log('Original: ', soup.prettify())
#Find article
contents = soup.find('div', {'class':['article_text']})
if not contents:
self.log('postprocess_html: article div not found!')
return soup
contents.extract()
#Find title
title = soup.find('h1')
if title:
contents.insert(0, title)
#Find article image
newstop = soup.find('div', {'class':['newstop']})
if newstop:
img = newstop.find('img')
if img:
imgDiv = Tag(soup, 'div')
imgDiv['class'] = 'article_img'
if img.has_key('width'):
del(img['width'])
if img.has_key('height'):
del(img['height'])
#find description
element = img.parent.nextSibling
img.extract()
imgDiv.insert(0, img)
while element:
if not isinstance(element, Tag):
continue
nextElement = element.nextSibling
if 'p' == element.name:
element.extract()
element['class'] = 'article_img_desc'
imgDiv.insert(len(imgDiv.contents), element)
element = nextElement
contents.insert(1, imgDiv)
#find article abstract
abstract = soup.find('p', {'class':['subhead']})
if abstract:
abstract['class'] = 'article_desc'
contents.insert(2, abstract)
#Find article authors
authorsDiv = soup.find('div', {'class':['autors']})
if authorsDiv:
authorsP = authorsDiv.find('p')
if authorsP:
authorsP['class'] = 'article_authors'
contents.insert(len(contents.contents), authorsP)
#Fix urls that use relative path
urls = contents.findAll('a');
if urls:
for url in urls:
if not url.has_key('href'):
continue
if '/' == url['href'][0]:
url['href'] = self.base_url + url['href']
body = soup.find('td', {'class':['second_content']})
if body:
body.replaceWith(contents)
self.log('Result: ', soup.prettify())
return soup
我面临的问题是,即使最简单的减少,我也无法使文章正文出现在结果.MOBI中 - 评论keep_only_tags
,remove_tags_after
,{{1}变量设置。我所得到的只是周围的横幅和导航,但没有文章文本。日志文件没有提到“postprocess_html:找不到文章div”,这看起来像文章文本丢失的主要信号。
当然,同一RSS的基于RSS的简单自定义新闻源工作得很好:所有文章正文都包括在内。
有关如何解决问题的任何建议?
答案 0 :(得分:0)
通过简单地评论postprocess_html
方法的整个主体来解决问题。