def parseList(self, response):
source = response.meta['source']
wx_account = response.meta['wx_account']
url = response.meta['url']
body = EncodeUtil.toUnicode(response.body)
# 判断被禁止 提示需要重启路由 清理cookie
if response.status == 302:
# 更新状态为更新失败
self.logDao.warn(u'您的访问过于频繁,重新拨号')
self.wxSourceDao.updateStatus(wx_account, 'updateFail')
# 获取Ip # 同时空线程30s
NetworkUtil.getNewIp()
TimerUtil.sleep(30)
else:
self.logDao.info(u'开始解析:' + wx_account)
# 进行解析
selector = Selector(text=body)
results = selector.xpath('//*[@id="main"]/div[4]/ul/li')
self.logDao.info(u'列表长度:' + str(len(results)))
hasCatch = False
for result in results:
wx_name = result.xpath('//*[@id="sogou_vr_11002301_box_0"]/div/div[2]/p[1]/a/text()').extract_first()
wx_account_ = result.xpath('//p[@class="info"]/label/text()').extract_first()
wx_url = result.xpath('//p[@class="tit"]/a/@href').extract_first()
if wx_account_ == wx_account:
self.logDao.info(u'成功抓取:' + wx_account_)
self.wxSourceDao.updateSource(wx_account, wx_name, wx_url, 'last')
hasCatch = True
break
if not hasCatch:
self.logDao.info(u'没有抓到:' + wx_account_)
self.wxSourceDao.updateStatus(wx_account, 'none')
pass
def parseArticleList(self, response):
body = EncodeUtil.toUnicode(response.body)
selector = Selector(text=body)
source_url = response.meta['source_url']
print source_url
title = selector.xpath('//title/text()').extract_first('').strip(u' ')
isN = u"请输入验证码" == title
if isN or response.status == 302:
self.logDao.info(u'访问过多被禁止,重新拨号')
# 获取Ip # 同时空线程30s
NetworkUtil.getNewIp()
TimerUtil.sleep(50)
NetworkUtil.openWebbrowser(source_url)
else:
source = response.meta['source']
wx_account = response.meta['wx_account']
wx_account_id = response.meta['wx_account_id']
self.logDao.info(u'开始解析列表:' + wx_account)
# 进行解析
articleJS = selector.xpath('//script/text()').extract()
for js in articleJS:
if 'var msgList = ' in js:
p8 = re.compile('var\s*msgList\s*=.*;')
matchList = p8.findall(js)
for match in matchList:
match = match.lstrip('var msgList = ').rstrip(';')
# 格式化
articles = demjson.decode(match) or {}
articles = articles['list'] or []
self.logDao.info(u'匹配到文章列表' + wx_account)
for article in articles:
app_msg_ext_info = article.get(
'app_msg_ext_info') or {}
desc = app_msg_ext_info.get('digest') or ''
title = app_msg_ext_info.get('title') or ''
# 如果存在则不抓取
if self.checkDao.checkExist(title, wx_account, 1):
self.logDao.info(u'已经存在' + wx_account + ':' +
title)
continue
detailUrl = app_msg_ext_info['content_url'] or ''
if not detailUrl:
continue
detailUrl = "http://mp.weixin.qq.com" + detailUrl
detailUrl = detailUrl.replace("amp;", "")
self.logDao.info(u'抓取' + wx_account + ':' + title +
':' + detailUrl)
yield scrapy.Request(url=detailUrl,
meta={
'request_type':
'weixin_detail',
'wx_account': wx_account,
"source": source,
"title": title,
'wx_account_id':
wx_account_id,
"source_url": detailUrl
},
callback=self.parseArticle)
def parseArticle(self, response):
body = EncodeUtil.toUnicode(response.body)
selector = Selector(text=body)
title = selector.xpath('//title/text()').extract_first('').strip(u' ')
source_url = response.meta['source_url']
wx_account = response.meta['wx_account']
isN = u"请输入验证码" == title
if isN or response.status == 302:
self.logDao.info(u'访问过多被禁止,重新拨号')
# 存起来
self.brokenAccounts.append(wx_account)
# 获取Ip # 同时空线程30s
NetworkUtil.getNewIp()
TimerUtil.sleep(80)
NetworkUtil.openWebbrowser(source_url)
else:
title = response.meta['title']
source_url = response.meta['source_url']
wx_account_id = response.meta['wx_account_id']
self.logDao.info(u'开始解析文章' + wx_account + ':' + title + ':' +
source_url)
self.logDao.info(u'开始解析文章：' + source_url)
# 进行解析
post_date = selector.xpath(
'//*[@id="post-date"]/text()').extract_first('')
try:
post_date = time.strftime("%Y-%m-%d %H:%M:%S",
time.strptime(post_date, "%Y-%m-%d"))
except Exception:
pass
styles = selector.xpath('//style/text()').extract()
styles = CssUtil.compressCss(styles).replace('\'', '"').replace(
'\\', '\\\\')
styles = CssUtil.clearUrl(styles)
styles = CssUtil.clearBackgroundColor(styles, ['#f3f3f3'])
post_user = selector.xpath(
'//*[@id="post-user"]/text()').extract_first('')
content_html = selector.xpath('//*[@id="js_content"]')
if not len(content_html):
self.logDao.info(u'不存在内容：' + source_url)
return
# 去除内部不需要的标签
content_items = content_html.xpath('*')
if not len(content_items):
self.logDao.info(u'不存在内容：' + source_url)
return
# content_items_new = []
# for item in content_items:
# itemStr = item.extract()
# if u'订阅微信' in itemStr:
# continue
# content_items_new.append(item)
# content_items = content_items_new
# 得到纯文本
content_txt = []
for item in content_items:
# 文本
allTxt = item.xpath('.//text()').extract()
allTxt = ''.join(allTxt).replace('\t', '')
# 加入
content_txt.append(allTxt)
content_txt = '\n'.join(content_txt)
# 组装新的内容标签
outHtml = """<div class="rich_media_content " id="js_content">${++content++}</div>"""
content_items = content_items.extract()
content_items = ''.join(content_items)
content_html = outHtml.replace('${++content++}', content_items)
selector = Selector(text=content_html)
# 解析文档中的所有图片url，然后替换成标识
image_urls = []
imgs = selector.xpath('descendant::img')
for img in imgs:
# 图片可能放在src 或者data-src
image_url = img.xpath('@src | @data-src').extract_first('')
if image_url and image_url.startswith('http'):
self.logDao.info(u'得到图片：' + image_url)
image_urls.append({
'url': image_url,
})
self.logDao.info(wx_account + u'得到文章：' + title + ":" + post_date +
':' + post_user)
self.logDao.info(u'得到文章：' + source_url)
# 得到hashCode1
hash_code = self.checkDao.getHashCode(title, wx_account, 1)
self.saveFile(hash_code, body)
# 去除 image 的 alt title
selector = Selector(text=content_html)
imgAltTitles = selector.xpath('//img/@alt|//img/@title').extract()
# 处理提示块img的 alt title, 关注//img/@alt|//img/@title
for imgAltTitle in imgAltTitles:
if imgAltTitle.strip(' '):
content_html = content_html.replace(imgAltTitle, '')
contentItem = ContentItem()
contentItem['content_txt'] = content_txt
contentItem['image_urls'] = image_urls
contentItem['title'] = title
contentItem['source_url'] = source_url
contentItem['post_date'] = post_date
contentItem['sub_channel'] = ''
contentItem['post_user'] = post_user
contentItem['tags'] = ''
contentItem['styles'] = styles
contentItem['content_html'] = content_html
contentItem['hash_code'] = hash_code
contentItem['info_type'] = 1
contentItem['src_source_id'] = 1
contentItem['src_account_id'] = wx_account_id
contentItem['src_channel'] = '微信公众号'
contentItem['src_ref'] = ''
contentItem['wx_account'] = wx_account
return contentItem