def spider_error(self, failure, response, spider):
message = 'Error: {0}'.format(failure.getErrorMessage())
ei_item = ErrorInfoItem()
ei_item['time'] = now_string()
ei_item['url'] = response.url
ei_item['error_level'] = "E"
ei_item['error_type'] = "E1000"
ei_item['description'] = message
spider.logger.error(message)
return ei_item
def spider_closed(self, spider):
now = now_string()
message = 'bdbk spider end at: {0}'.format(now)
ei_item = ErrorInfoItem()
ei_item['time'] = now
ei_item['url'] = spider.start_page
ei_item['error_level'] = "I"
ei_item['error_type'] = "I2"
ei_item['description'] = message
spider.logger.info(message)
return ei_item
def parse_person(self, response):
url = response.url.split('?')[0]
if self.check_visited(url):
return
kwlist = response.xpath('//meta[@name="keywords"]/@content').extract()
if len(kwlist) == 0:
self.redis_client_person.set(url, 1)
return
keywords = kwlist[0].encode('utf-8', 'ignore')
'''
# the 'keywords' meta must contains '人物'
if keywords.find('人物') == -1:
self.redis_client_person.set(url, 1)
return
'''
description = response.xpath('//meta[@name="description"]/@content').extract()[0].encode('utf-8', 'ignore')
page_title = response.xpath('//h1/text()').extract()[0].encode('utf-8', 'ignore')
# get person tags (人物标签)
person_tags = list()
categories = dict()
is_person = False
for sel in response.xpath('//span[@class="taglist"]'):
tag = sel.xpath('text()').extract()[0]
tag = re.sub(r'[\r\n]*', '', tag).encode('utf-8', 'ignore')
if len(tag) == 0:
continue
if tag in self.ignore_tags:
message = 'In ignore list. name: {0}, tag: {1}'.format(page_title, tag)
ei_item = ErrorInfoItem()
ei_item['time'] = now_string()
ei_item['url'] = url
ei_item['error_level'] = "W"
ei_item['error_type'] = 'W1'
ei_item['description'] = message
yield ei_item
self.logger.warning(message)
self.redis_client_person.set(url, 1)
return
if tag.find('人物') != -1:
is_person = True
person_tags.append(tag)
# save to redis
category_cnt = self.redis_client.get(tag)
if str(category_cnt) == 'None':
category_cnt = 1
else:
category_cnt = int(category_cnt) + 1
self.redis_client.set(tag, category_cnt)
categories[tag] = category_cnt
# if tags do not contains |人物|, just follow link
if is_person == False and self.follow_link == True:
self.redis_client_person.set(url, 1)
# follow link that which url contains |view|(view/subview)
for sel in response.xpath('//a[contains(@href, "view")]'):
url = response.urljoin(sel.xpath('@href').extract()[0].split('?')[0])
if self.check_visited(url):
return
request = scrapy.Request(url, callback = self.parse_person)
yield request
return
person_item = PersonItem()
person_item['name'] = page_title
person_item['url'] = url
person_item['description'] = description
person_item['tags'] = person_tags
person_item['keywords'] = keywords
summary_pic = response.xpath('//div[@class="summary-pic"]/a/img/@src').extract()
if len(summary_pic) > 0:
summary_pic = summary_pic[0].split('/')[-1].split('.')[0]
else:
summary_pic = ''
person_item['summary_pic'] = summary_pic
# for the data pipeline
yield person_item
yield categories
# crawling image gallery (图册)
# album list
album_list = response.xpath('//script/text()').re(r'AlbumList\({.*[\n\t]*.*[\n\t]*.*[\n\t]*.*')
albums = list()
if len(album_list) > 0:
album_list = album_list[0]
album_list = re.sub(r'[\r\n\t]*', '', album_list)
album_lemma_id = re.findall(r'lemmaId:"([\d]+)"', album_list)[0]
album_sublemma_id = re.findall(r'subLemmaId:"([\d]+)"', album_list)[0]
album_data_json = re.sub(r'AlbumList.*data:', '', album_list)
try:
album_data_dict = json.loads(album_data_json)
i = 0
for d in album_data_dict:
if isinstance(album_data_dict, list):
cover_pic = d["coverpic"]
album_desc= d["desc"]
album_total= d["total"]
album_url = '/picture/{0}/{1}/{2}/{3}'.format(album_lemma_id, album_sublemma_id, i, cover_pic)
i += 1
else:
cover_pic = album_data_dict[d]["coverpic"]
album_desc= album_data_dict[d]["desc"]
album_total= album_data_dict[d]["total"]
album_url = '/picture/{0}/{1}/{2}/{3}'.format(album_lemma_id, album_sublemma_id, d, cover_pic)
album_url = response.urljoin(album_url)
# build album_item
album_item = AlbumItem()
album_item['url'] = album_url
album_item['description'] = album_desc.encode('utf8', 'ignore')
album_item['total'] = album_total
album_item['cover_pic'] = cover_pic
album_item['person_name'] = person_item['name']
album_item['person_url'] = person_item['url']
albums.append(album_item)
except Exception, e:
self.logger.error('json parse album list info error. url: %s, err: %r', response.url, e)
try:
r = re.compile('albums:.*,[\r\n\s]*lemmaId:')
for s in response.xpath('//script/text()').extract():
match = re.search(r, s)
if match:
album_info_str = match.group()
album_info_str = re.sub(r',[\r\n\s]*lemmaId:', '', album_info_str)
album_info_str = "{%s}" % album_info_str.replace('albums', '"albums"')
break
except Exception, e:
self.logger.error('get album info json error. url: %s, err: %r', response.url, e)
return
if album_info_str == None:
message = 'Album not found. person_name: {0}, person_url: {1}'.format(person_info['name'], person_info['url'])
ei_item = ErrorInfoItem()
ei_item['time'] = now_string()
ei_item['url'] = response.url
ei_item['error_level'] = "E"
ei_item['error_type'] = "E1"
ei_item['description'] = message
yield ei_item
self.logger.warning('{%s}. url: %s', message, response.url)
return
album_info_dic = None
try:
album_info_dic = json.loads(album_info_str)
album_info_dic = album_info_dic['albums']
except Exception, e:
message = 'json.loads album info error. url: {0}, json: {1}, err: {2}'.format(response.url, album_info_str, e)
ei_item = ErrorInfoItem()
