class SearchSpider(scrapy.Spider):
name = 'search'
allowed_domains = ['weibo.com']
settings = get_project_settings()
keyword_list = settings.get('KEYWORD_LIST')
if not isinstance(keyword_list, list):
if not os.path.isabs(keyword_list):
keyword_list = os.getcwd() + os.sep + keyword_list
if not os.path.isfile(keyword_list):
print('不存在%s文件' % keyword_list)
sys.exit()
keyword_list = util.get_keyword_list(keyword_list)
for i, keyword in enumerate(keyword_list):
if len(keyword) > 2 and keyword[0] == '#' and keyword[-1] == '#':
keyword_list[i] = '%23' + keyword[1:-1] + '%23'
weibo_type = util.convert_weibo_type(settings.get('WEIBO_TYPE'))
contain_type = util.convert_contain_type(settings.get('CONTAIN_TYPE'))
regions = util.get_regions(settings.get('REGION'))
base_url = 'https://s.weibo.com'
start_date = settings.get('START_DATE',
datetime.now().strftime('%Y-%m-%d'))
end_date = settings.get('END_DATE', datetime.now().strftime('%Y-%m-%d'))
further_threshold = settings.get('FURTHER_THRESHOLD', 46)
mongo_error = False
pymongo_error = False
mysql_error = False
pymysql_error = False
# 连接mongodb
MAX_COUNT = 3000
isRunning = True
client = pymongo.MongoClient(host='124.70.84.12', port=27017)
db = client['weibo_keyword_epidemic']
collection = db.get_collection(str(start_date).split(' ')[0])
def start_requests(self):
start_date = datetime.strptime(self.start_date, '%Y-%m-%d')
end_date = datetime.strptime(self.end_date,
'%Y-%m-%d') + timedelta(days=1)
start_str = start_date.strftime('%Y-%m-%d') + '-0'
end_str = end_date.strftime('%Y-%m-%d') + '-0'
for keyword in self.keyword_list:
if not self.settings.get('REGION') or '全部' in self.settings.get(
'REGION'):
base_url = 'https://s.weibo.com/weibo?q=%s' % keyword
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}'.format(start_str, end_str)
if not self.isRunning:
os.abort()
yield scrapy.Request(url=url,
callback=self.parse,
meta={
'base_url': base_url,
'keyword': keyword
})
if not self.isRunning:
os.abort()
else:
for region in self.regions.values():
base_url = (
'https://s.weibo.com/weibo?q={}®ion=custom:{}:1000'
).format(keyword, region['code'])
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}'.format(start_str, end_str)
# 获取一个省的搜索结果
yield scrapy.Request(url=url,
callback=self.parse,
meta={
'base_url': base_url,
'keyword': keyword,
'province': region
})
if not self.isRunning:
os.abort()
def check_environment(self):
"""判断配置要求的软件是否已安装"""
if self.pymongo_error:
print('系统中可能没有安装pymongo库，请先运行 pip install pymongo ，再运行程序')
raise CloseSpider()
if self.mongo_error:
print('系统中可能没有安装或启动MongoDB数据库，请先根据系统环境安装或启动MongoDB，再运行程序')
raise CloseSpider()
if self.pymysql_error:
print('系统中可能没有安装pymysql库，请先运行 pip install pymysql ，再运行程序')
raise CloseSpider()
if self.mysql_error:
print('系统中可能没有安装或正确配置MySQL数据库，请先根据系统环境安装或配置MySQL，再运行程序')
raise CloseSpider()
def parse(self, response):
if self.collection.count() > self.MAX_COUNT:
self.isRunning = False
os.abort()
base_url = response.meta.get('base_url')
keyword = response.meta.get('keyword')
province = response.meta.get('province')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < self.further_threshold:
'''解析当前页面'''
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
start_date = datetime.strptime(self.start_date, '%Y-%m-%d')
end_date = datetime.strptime(self.end_date, '%Y-%m-%d')
while start_date <= end_date:
start_str = start_date.strftime('%Y-%m-%d') + '-0'
start_date = start_date + timedelta(days=1)
end_str = start_date.strftime('%Y-%m-%d') + '-0'
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_str, end_str)
# 获取一天的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_by_day,
meta={
'base_url': base_url,
'keyword': keyword,
'province': province,
'date': start_str[:-2]
})
def parse_by_day(self, response):
if self.collection.count() > self.MAX_COUNT:
self.isRunning = False
os.abort()
"""以天为单位筛选"""
base_url = response.meta.get('base_url')
keyword = response.meta.get('keyword')
province = response.meta.get('province')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
date = response.meta.get('date')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < self.further_threshold:
'''解析当前页面'''
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
start_date_str = date + '-0'
start_date = datetime.strptime(start_date_str, '%Y-%m-%d-%H')
for i in range(1, 25):
start_str = start_date.strftime('%Y-%m-%d-X%H').replace(
'X0', 'X').replace('X', '')
start_date = start_date + timedelta(hours=1)
end_str = start_date.strftime('%Y-%m-%d-X%H').replace(
'X0', 'X').replace('X', '')
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_str, end_str)
# 获取一小时的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_by_hour_province
if province else self.parse_by_hour,
meta={
'base_url': base_url,
'keyword': keyword,
'province': province,
'start_time': start_str,
'end_time': end_str
})
def parse_by_hour(self, response):
if self.collection.count() > self.MAX_COUNT:
self.isRunning = False
os.abort()
"""以小时为单位筛选"""
keyword = response.meta.get('keyword')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
start_time = response.meta.get('start_time')
end_time = response.meta.get('end_time')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < self.further_threshold:
'''解析当前页面'''
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
for region in self.regions.values():
url = ('https://s.weibo.com/weibo?q={}®ion=custom:{}:1000'
).format(keyword, region['code'])
url += self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_time, end_time)
# 获取一小时一个省的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_by_hour_province,
meta={
'keyword': keyword,
'start_time': start_time,
'end_time': end_time,
'province': region
})
def parse_by_hour_province(self, response):
if self.collection.count() > self.MAX_COUNT:
self.isRunning = False
os.abort()
"""以小时和直辖市/省为单位筛选"""
keyword = response.meta.get('keyword')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
start_time = response.meta.get('start_time')
end_time = response.meta.get('end_time')
province = response.meta.get('province')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < self.further_threshold:
# 解析当前页面
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
for city in province['city'].values():
url = ('https://s.weibo.com/weibo?q={}®ion=custom:{}:{}'
).format(keyword, province['code'], city)
url += self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_time, end_time)
# 获取一小时一个城市的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_page,
meta={
'keyword': keyword,
'start_time': start_time,
'end_time': end_time,
'province': province,
'city': city
})
def parse_page(self, response):
if self.collection.count() > self.MAX_COUNT:
self.isRunning = False
os.abort()
"""解析一页搜索结果的信息"""
keyword = response.meta.get('keyword')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
if is_empty:
print('当前页面搜索结果为空')
else:
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
def get_article_url(self, selector):
"""获取微博头条文章url"""
article_url = ''
text = selector.xpath('string(.)').extract_first().replace(
'\u200b', '').replace('\ue627', '').replace('\n',
'').replace(' ', '')
if text.startswith('发布了头条文章'):
urls = selector.xpath('.//a')
for url in urls:
if url.xpath(
'i[@class="wbicon"]/text()').extract_first() == 'O':
if url.xpath('@href').extract_first() and url.xpath(
'@href').extract_first().startswith('http://t.cn'):
article_url = url.xpath('@href').extract_first()
break
return article_url
def get_location(self, selector):
"""获取微博发布位置"""
a_list = selector.xpath('.//a')
location = ''
for a in a_list:
if a.xpath('./i[@class="wbicon"]') and a.xpath(
'./i[@class="wbicon"]/text()').extract_first() == '2':
location = a.xpath('string(.)').extract_first()[1:]
break
return location
def get_at_users(self, selector):
"""获取微博中@的用户昵称"""
a_list = selector.xpath('.//a')
at_users = ''
at_list = []
for a in a_list:
if len(unquote(a.xpath('@href').extract_first())) > 14 and len(
a.xpath('string(.)').extract_first()) > 1:
if unquote(a.xpath('@href').extract_first())[14:] == a.xpath(
'string(.)').extract_first()[1:]:
at_user = a.xpath('string(.)').extract_first()[1:]
if at_user not in at_list:
at_list.append(at_user)
if at_list:
at_users = ','.join(at_list)
return at_users
def get_topics(self, selector):
"""获取参与的微博话题"""
a_list = selector.xpath('.//a')
topics = ''
topic_list = []
for a in a_list:
text = a.xpath('string(.)').extract_first()
if len(text) > 2 and text[0] == '#' and text[-1] == '#':
if text[1:-1] not in topic_list:
topic_list.append(text[1:-1])
if topic_list:
topics = ','.join(topic_list)
return topics
def parse_weibo(self, response):
"""解析网页中的微博信息"""
keyword = response.meta.get('keyword')
for sel in response.xpath("//div[@class='card-wrap']"):
info = sel.xpath(
"div[@class='card']/div[@class='card-feed']/div[@class='content']/div[@class='info']"
)
if info:
weibo = WeiboItem()
weibo['id'] = sel.xpath('@mid').extract_first()
weibo['bid'] = sel.xpath(
'(.//p[@class="from"])[last()]/a[1]/@href').extract_first(
).split('/')[-1].split('?')[0]
weibo['user_id'] = info[0].xpath(
'div[2]/a/@href').extract_first().split('?')[0].split(
'/')[-1]
weibo['screen_name'] = info[0].xpath(
'div[2]/a/@nick-name').extract_first()
txt_sel = sel.xpath('.//p[@class="txt"]')[0]
retweet_sel = sel.xpath('.//div[@class="card-comment"]')
retweet_txt_sel = ''
if retweet_sel and retweet_sel[0].xpath('.//p[@class="txt"]'):
retweet_txt_sel = retweet_sel[0].xpath(
'.//p[@class="txt"]')[0]
content_full = sel.xpath(
'.//p[@node-type="feed_list_content_full"]')
is_long_weibo = False
is_long_retweet = False
if content_full:
if not retweet_sel:
txt_sel = content_full[0]
is_long_weibo = True
elif len(content_full) == 2:
txt_sel = content_full[0]
retweet_txt_sel = content_full[1]
is_long_weibo = True
is_long_retweet = True
elif retweet_sel[0].xpath(
'.//p[@node-type="feed_list_content_full"]'):
retweet_txt_sel = retweet_sel[0].xpath(
'.//p[@node-type="feed_list_content_full"]')[0]
is_long_retweet = True
else:
txt_sel = content_full[0]
is_long_weibo = True
weibo['text'] = txt_sel.xpath(
'string(.)').extract_first().replace('\u200b', '').replace(
'\ue627', '')
weibo['article_url'] = self.get_article_url(txt_sel)
weibo['location'] = self.get_location(txt_sel)
if weibo['location']:
weibo['text'] = weibo['text'].replace(
'2' + weibo['location'], '')
weibo['text'] = weibo['text'][2:].replace(' ', '')
if is_long_weibo:
weibo['text'] = weibo['text'][:-6]
weibo['at_users'] = self.get_at_users(txt_sel)
weibo['topics'] = self.get_topics(txt_sel)
reposts_count = sel.xpath(
'.//a[@action-type="feed_list_forward"]/text()'
).extract_first()
try:
reposts_count = re.findall(r'\d+.*', reposts_count)
except TypeError:
print('cookie无效或已过期，请按照'
'https://github.com/dataabc/weibo-search#如何获取cookie'
' 获取cookie')
raise CloseSpider()
weibo['reposts_count'] = reposts_count[
0] if reposts_count else '0'
comments_count = sel.xpath(
'.//a[@action-type="feed_list_comment"]/text()'
).extract_first()
comments_count = re.findall(r'\d+.*', comments_count)
weibo['comments_count'] = comments_count[
0] if comments_count else '0'
attitudes_count = sel.xpath(
'(.//a[@action-type="feed_list_like"])[last()]/em/text()'
).extract_first()
weibo['attitudes_count'] = (attitudes_count
if attitudes_count else '0')
created_at = sel.xpath(
'(.//p[@class="from"])[last()]/a[1]/text()').extract_first(
).replace(' ', '').replace('\n', '').split('前')[0]
weibo['created_at'] = util.standardize_date(created_at)
source = sel.xpath('(.//p[@class="from"])[last()]/a[2]/text()'
).extract_first()
weibo['source'] = source if source else ''
pics = ''
is_exist_pic = sel.xpath(
'.//div[@class="media media-piclist"]')
if is_exist_pic:
pics = is_exist_pic[0].xpath('ul[1]/li/img/@src').extract()
pics = [pic[2:] for pic in pics]
pics = [
re.sub(r'/.*?/', '/large/', pic, 1) for pic in pics
]
pics = ['http://' + pic for pic in pics]
video_url = ''
is_exist_video = sel.xpath(
'.//div[@class="thumbnail"]/a/@action-data')
if is_exist_video:
video_url = is_exist_video.extract_first()
video_url = unquote(
str(video_url)).split('video_src=//')[-1]
video_url = 'http://' + video_url
if not retweet_sel:
weibo['pics'] = pics
weibo['video_url'] = video_url
else:
weibo['pics'] = ''
weibo['video_url'] = ''
weibo['retweet_id'] = ''
if retweet_sel and retweet_sel[0].xpath(
'.//div[@node-type="feed_list_forwardContent"]/a[1]'):
retweet = WeiboItem()
retweet['id'] = retweet_sel[0].xpath(
'.//a[@action-type="feed_list_like"]/@action-data'
).extract_first()[4:]
retweet['bid'] = retweet_sel[0].xpath(
'.//p[@class="from"]/a/@href').extract_first().split(
'/')[-1].split('?')[0]
info = retweet_sel[0].xpath(
'.//div[@node-type="feed_list_forwardContent"]/a[1]'
)[0]
retweet['user_id'] = info.xpath(
'@href').extract_first().split('/')[-1]
retweet['screen_name'] = info.xpath(
'@nick-name').extract_first()
retweet['text'] = retweet_txt_sel.xpath(
'string(.)').extract_first().replace('\u200b',
'').replace(
'\ue627', '')
retweet['article_url'] = self.get_article_url(
retweet_txt_sel)
retweet['location'] = self.get_location(retweet_txt_sel)
if retweet['location']:
retweet['text'] = retweet['text'].replace(
'2' + retweet['location'], '')
retweet['text'] = retweet['text'][2:].replace(' ', '')
if is_long_retweet:
retweet['text'] = retweet['text'][:-6]
retweet['at_users'] = self.get_at_users(retweet_txt_sel)
retweet['topics'] = self.get_topics(retweet_txt_sel)
reposts_count = retweet_sel[0].xpath(
'.//ul[@class="act s-fr"]/li/a[1]/text()'
).extract_first()
reposts_count = re.findall(r'\d+.*', reposts_count)
retweet['reposts_count'] = reposts_count[
0] if reposts_count else '0'
comments_count = retweet_sel[0].xpath(
'.//ul[@class="act s-fr"]/li[2]/a[1]/text()'
).extract_first()
comments_count = re.findall(r'\d+.*', comments_count)
retweet['comments_count'] = comments_count[
0] if comments_count else '0'
attitudes_count = retweet_sel[0].xpath(
'.//a[@action-type="feed_list_like"]/em/text()'
).extract_first()
retweet['attitudes_count'] = (attitudes_count
if attitudes_count else '0')
created_at = retweet_sel[0].xpath(
'.//p[@class="from"]/a[1]/text()').extract_first(
).replace(' ', '').replace('\n', '').split('前')[0]
retweet['created_at'] = util.standardize_date(created_at)
source = retweet_sel[0].xpath(
'.//p[@class="from"]/a[2]/text()').extract_first()
retweet['source'] = source if source else ''
retweet['pics'] = pics
retweet['video_url'] = video_url
retweet['retweet_id'] = ''
yield {'weibo': retweet, 'keyword': keyword}
weibo['retweet_id'] = retweet['id']
print(weibo)
yield {'weibo': weibo, 'keyword': keyword}
class SearchSpider(scrapy.Spider):
name = 'search'
allowed_domains = ['weibo.com']
settings = get_project_settings()
keyword_list = settings.get('KEYWORD_LIST')
if not isinstance(keyword_list, list):
if not os.path.isabs(keyword_list):
keyword_list = os.getcwd() + os.sep + keyword_list
if not os.path.isfile(keyword_list):
print('不存在%s文件' % keyword_list)
sys.exit()
keyword_list = util.get_keyword_list(keyword_list)
for i, keyword in enumerate(keyword_list):
if len(keyword) > 2 and keyword[0] == '#' and keyword[-1] == '#':
keyword_list[i] = '%23' + keyword[1:-1] + '%23'
weibo_type = util.convert_weibo_type(settings.get('WEIBO_TYPE'))
contain_type = util.convert_contain_type(settings.get('CONTAIN_TYPE'))
regions = util.get_regions(settings.get('REGION'))
base_url = 'https://s.weibo.com'
start_date = settings.get('START_DATE',
datetime.now().strftime('%Y-%m-%d'))
end_date = settings.get('END_DATE', datetime.now().strftime('%Y-%m-%d'))
login_cookies = util.load_cookies('./cookies/cookies.json')
mongo_error = False
pymongo_error = False
mysql_error = False
pymysql_error = False
#self.cookies
# cookies=[]
# DEFAULT_REQUEST_HEADERS=settings.get('DEFAULT_REQUEST_HEADERS')
# for name_value in DEFAULT_REQUEST_HEADERS['cookie'].split(' '):
# name,value=name_value.split('=')
# cookies.append({'name':name,'value':value})
# print(cookies)
def start_requests(self):
start_date = datetime.strptime(self.start_date, '%Y-%m-%d')
end_date = datetime.strptime(self.end_date,
'%Y-%m-%d') + timedelta(days=1)
start_str = start_date.strftime('%Y-%m-%d') + '-0'
end_str = end_date.strftime('%Y-%m-%d') + '-0'
for keyword in self.keyword_list:
if not self.settings.get('REGION') or '全部' in self.settings.get(
'REGION'):
base_url = 'https://s.weibo.com/weibo?q=%s' % keyword
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}'.format(start_str, end_str)
yield scrapy.Request(url=url,
callback=self.parse,
meta={
'base_url': base_url,
'keyword': keyword,
'url': url
})
else:
for region in self.regions.values():
base_url = (
'https://s.weibo.com/weibo?q={}®ion=custom:{}:1000'
).format(keyword, region['code'])
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}'.format(start_str, end_str)
# 获取一个省的搜索结果
yield scrapy.Request(url=url,
callback=self.parse,
meta={
'base_url': base_url,
'keyword': keyword,
'province': region,
'url': url
})
def check_environment(self):
"""判断配置要求的软件是否已安装"""
if self.pymongo_error:
print('系统中可能没有安装pymongo库，请先运行 pip install pymongo ，再运行程序')
raise CloseSpider()
if self.mongo_error:
print('系统中可能没有安装或启动MongoDB数据库，请先根据系统环境安装或启动MongoDB，再运行程序')
raise CloseSpider()
if self.pymysql_error:
print('系统中可能没有安装pymysql库，请先运行 pip install pymysql ，再运行程序')
raise CloseSpider()
if self.mysql_error:
print('系统中可能没有安装或正确配置MySQL数据库，请先根据系统环境安装或配置MySQL，再运行程序')
raise CloseSpider()
def parse(self, response):
base_url = response.meta.get('base_url')
keyword = response.meta.get('keyword')
province = response.meta.get('province')
url = response.meta.get('url')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < 50:
# 解析当前页面
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
start_date = datetime.strptime(self.start_date, '%Y-%m-%d')
end_date = datetime.strptime(self.end_date, '%Y-%m-%d')
while start_date <= end_date:
start_str = start_date.strftime('%Y-%m-%d') + '-0'
start_date = start_date + timedelta(days=1)
end_str = start_date.strftime('%Y-%m-%d') + '-0'
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_str, end_str)
# 获取一天的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_by_day,
meta={
'base_url': base_url,
'keyword': keyword,
'province': province,
'date': start_str[:-2]
})
def parse_by_day(self, response):
"""以天为单位筛选"""
base_url = response.meta.get('base_url')
keyword = response.meta.get('keyword')
province = response.meta.get('province')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
date = response.meta.get('date')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < 50:
# 解析当前页面
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
start_date_str = date + '-0'
start_date = datetime.strptime(start_date_str, '%Y-%m-%d-%H')
for i in range(1, 25):
start_str = start_date.strftime('%Y-%m-%d-X%H').replace(
'X0', 'X').replace('X', '')
start_date = start_date + timedelta(hours=1)
end_str = start_date.strftime('%Y-%m-%d-X%H').replace(
'X0', 'X').replace('X', '')
url = base_url + self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_str, end_str)
# 获取一小时的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_by_hour_province
if province else self.parse_by_hour,
meta={
'base_url': base_url,
'keyword': keyword,
'province': province,
'start_time': start_str,
'end_time': end_str
})
def parse_by_hour(self, response):
"""以小时为单位筛选"""
keyword = response.meta.get('keyword')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
start_time = response.meta.get('start_time')
end_time = response.meta.get('end_time')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < 50:
# 解析当前页面
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
for region in self.regions.values():
url = ('https://s.weibo.com/weibo?q={}®ion=custom:{}:1000'
).format(keyword, region['code'])
url += self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_time, end_time)
# 获取一小时一个省的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_by_hour_province,
meta={
'keyword': keyword,
'start_time': start_time,
'end_time': end_time,
'province': region
})
def parse_by_hour_province(self, response):
"""以小时和直辖市/省为单位筛选"""
keyword = response.meta.get('keyword')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
start_time = response.meta.get('start_time')
end_time = response.meta.get('end_time')
province = response.meta.get('province')
page_count = len(response.xpath('//ul[@class="s-scroll"]/li'))
if is_empty:
print('当前页面搜索结果为空')
elif page_count < 50:
# 解析当前页面
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
else:
for city in province['city'].values():
url = ('https://s.weibo.com/weibo?q={}®ion=custom:{}:{}'
).format(keyword, province['code'], city)
url += self.weibo_type
url += self.contain_type
url += '×cope=custom:{}:{}&page=1'.format(
start_time, end_time)
# 获取一小时一个城市的搜索结果
yield scrapy.Request(url=url,
callback=self.parse_page,
meta={
'keyword': keyword,
'start_time': start_time,
'end_time': end_time,
'province': province,
'city': city
})
def parse_page(self, response):
"""解析一页搜索结果的信息"""
keyword = response.meta.get('keyword')
is_empty = response.xpath(
'//div[@class="card card-no-result s-pt20b40"]')
if is_empty:
print('当前页面搜索结果为空')
else:
for weibo in self.parse_weibo(response):
self.check_environment()
yield weibo
next_url = response.xpath(
'//a[@class="next"]/@href').extract_first()
if next_url:
next_url = self.base_url + next_url
yield scrapy.Request(url=next_url,
callback=self.parse_page,
meta={'keyword': keyword})
def get_article_url(self, selector):
"""获取微博头条文章url"""
article_url = ''
text = selector.xpath('string(.)').extract_first().replace(
'\u200b', '').replace('\ue627', '').replace('\n',
'').replace(' ', '')
if text.startswith('发布了头条文章'):
urls = selector.xpath('.//a')
for url in urls:
if url.xpath(
'i[@class="wbicon"]/text()').extract_first() == 'O':
if url.xpath('@href').extract_first() and url.xpath(
'@href').extract_first().startswith('http://t.cn'):
article_url = url.xpath('@href').extract_first()
break
return article_url
def get_location(self, selector):
"""获取微博发布位置"""
a_list = selector.xpath('.//a')
location = ''
for a in a_list:
if a.xpath('./i[@class="wbicon"]') and a.xpath(
'./i[@class="wbicon"]/text()').extract_first() == '2':
location = a.xpath('string(.)').extract_first()[1:]
break
return location
def get_at_users(self, selector):
"""获取微博中@的用户昵称"""
a_list = selector.xpath('.//a')
at_users = ''
at_list = []
for a in a_list:
if len(unquote(a.xpath('@href').extract_first())) > 14 and len(
a.xpath('string(.)').extract_first()) > 1:
if unquote(a.xpath('@href').extract_first())[14:] == a.xpath(
'string(.)').extract_first()[1:]:
at_user = a.xpath('string(.)').extract_first()[1:]
if at_user not in at_list:
at_list.append(at_user)
if at_list:
at_users = ','.join(at_list)
return at_users
def get_topics(self, selector):
"""获取参与的微博话题"""
a_list = selector.xpath('.//a')
topics = ''
topic_list = []
for a in a_list:
text = a.xpath('string(.)').extract_first()
if len(text) > 2 and text[0] == '#' and text[-1] == '#':
if text[1:-1] not in topic_list:
topic_list.append(text[1:-1])
if topic_list:
topics = ','.join(topic_list)
return topics
def parse_weibo(self, response):
"""解析网页中的微博信息"""
keyword = response.meta.get('keyword')
for sel in response.xpath("//div[@class='card-wrap']"):
info = sel.xpath(
"div[@class='card']/div[@class='card-feed']/div[@class='content']/div[@class='info']"
)
if info:
weibo = WeiboItem()
weibo['id'] = sel.xpath('@mid').extract_first()
weibo['bid'] = sel.xpath(
'(.//p[@class="from"])[last()]/a[1]/@href').extract_first(
).split('/')[-1].split('?')[0]
weibo['user_id'] = info[0].xpath(
'div[2]/a/@href').extract_first().split('?')[0].split(
'/')[-1]
weibo['screen_name'] = info[0].xpath(
'div[2]/a/@nick-name').extract_first()
# 用户类别
user_type = info[0].xpath('div[2]/a[2]/@title').extract_first()
if not user_type:
weibo['user_type'] = "其他"
else:
weibo['user_type'] = user_type
txt_sel = sel.xpath('.//p[@class="txt"]')[0]
retweet_sel = sel.xpath('.//div[@class="card-comment"]')
retweet_txt_sel = ''
if retweet_sel and retweet_sel[0].xpath('.//p[@class="txt"]'):
retweet_txt_sel = retweet_sel[0].xpath(
'.//p[@class="txt"]')[0]
content_full = sel.xpath(
'.//p[@node-type="feed_list_content_full"]')
is_long_weibo = False
is_long_retweet = False
if content_full:
if not retweet_sel:
txt_sel = content_full[0]
is_long_weibo = True
elif len(content_full) == 2:
txt_sel = content_full[0]
retweet_txt_sel = content_full[1]
is_long_weibo = True
is_long_retweet = True
elif retweet_sel[0].xpath(
'.//p[@node-type="feed_list_content_full"]'):
retweet_txt_sel = retweet_sel[0].xpath(
'.//p[@node-type="feed_list_content_full"]')[0]
is_long_retweet = True
else:
txt_sel = content_full[0]
is_long_weibo = True
weibo['text'] = txt_sel.xpath(
'string(.)').extract_first().replace('\u200b', '').replace(
'\ue627', '')
weibo['article_url'] = self.get_article_url(txt_sel)
weibo['location'] = self.get_location(txt_sel)
if weibo['location']:
weibo['text'] = weibo['text'].replace(
'2' + weibo['location'], '')
weibo['text'] = weibo['text'][2:].replace(' ', '')
if is_long_weibo:
weibo['text'] = weibo['text'][:-6]
weibo['at_users'] = self.get_at_users(txt_sel)
weibo['topics'] = self.get_topics(txt_sel)
reposts_count = sel.xpath(
'.//a[@action-type="feed_list_forward"]/text()'
).extract_first()
try:
reposts_count = re.findall(r'\d+.*', reposts_count)
except TypeError:
print('cookie无效或已过期，请按照'
'https://github.com/dataabc/weibo-search#如何获取cookie'
' 获取cookie')
raise CloseSpider()
weibo['reposts_count'] = reposts_count[
0] if reposts_count else '0'
comments_count = sel.xpath(
'.//a[@action-type="feed_list_comment"]/text()'
).extract_first()
comments_count = re.findall(r'\d+.*', comments_count)
weibo['comments_count'] = comments_count[
0] if comments_count else '0'
#Crawl Comments
if '万' in weibo['comments_count'] or int(
weibo['comments_count']) > 20:
#详细微博的url
comment_url = 'https://weibo.com/' + weibo[
'user_id'] + '/' + weibo['bid']
yield scrapy.Request(url=comment_url,
callback=self.parse_comments,
meta={
'weibo_id': weibo['id'],
'weibo_bid': weibo['bid'],
'weibo_user_id': weibo['user_id'],
'keyword': keyword
})
attitudes_count = sel.xpath(
'(.//a[@action-type="feed_list_like"])[last()]/em/text()'
).extract_first()
weibo['attitudes_count'] = (attitudes_count
if attitudes_count else '0')
created_at = sel.xpath(
'(.//p[@class="from"])[last()]/a[1]/text()').extract_first(
).replace(' ', '').replace('\n', '').split('前')[0]
weibo['created_at'] = util.standardize_date(created_at)
source = sel.xpath('(.//p[@class="from"])[last()]/a[2]/text()'
).extract_first()
weibo['source'] = source if source else ''
pics = ''
is_exist_pic = sel.xpath(
'.//div[@class="media media-piclist"]')
if is_exist_pic:
pics = is_exist_pic[0].xpath('ul[1]/li/img/@src').extract()
pics = [pic[2:] for pic in pics]
pics = [
re.sub(r'/.*?/', '/large/', pic, 1) for pic in pics
]
pics = ['http://' + pic for pic in pics]
video_url = ''
is_exist_video = sel.xpath(
'.//div[@class="thumbnail"]/a/@action-data')
if is_exist_video:
video_url = is_exist_video.extract_first()
video_url = unquote(
str(video_url)).split('video_src=//')[-1]
video_url = 'http://' + video_url
if not retweet_sel:
weibo['pics'] = pics
weibo['video_url'] = video_url
else:
weibo['pics'] = ''
weibo['video_url'] = ''
weibo['retweet_id'] = ''
if retweet_sel and retweet_sel[0].xpath(
'.//div[@node-type="feed_list_forwardContent"]/a[1]'):
retweet = WeiboItem()
retweet['id'] = retweet_sel[0].xpath(
'.//a[@action-type="feed_list_like"]/@action-data'
).extract_first()[4:]
retweet['bid'] = retweet_sel[0].xpath(
'.//p[@class="from"]/a/@href').extract_first().split(
'/')[-1].split('?')[0]
info = retweet_sel[0].xpath(
'.//div[@node-type="feed_list_forwardContent"]/a[1]'
)[0]
retweet['user_id'] = info.xpath(
'@href').extract_first().split('/')[-1]
retweet['screen_name'] = info.xpath(
'@nick-name').extract_first()
# 转发微博的用户类别
user_type = info.xpath('..//a[2]/@title').extract_first()
if not user_type:
retweet['user_type'] = "其他"
else:
retweet['user_type'] = user_type
retweet['text'] = retweet_txt_sel.xpath(
'string(.)').extract_first().replace('\u200b',
'').replace(
'\ue627', '')
retweet['article_url'] = self.get_article_url(
retweet_txt_sel)
retweet['location'] = self.get_location(retweet_txt_sel)
if retweet['location']:
retweet['text'] = retweet['text'].replace(
'2' + retweet['location'], '')
retweet['text'] = retweet['text'][2:].replace(' ', '')
if is_long_retweet:
retweet['text'] = retweet['text'][:-6]
retweet['at_users'] = self.get_at_users(retweet_txt_sel)
retweet['topics'] = self.get_topics(retweet_txt_sel)
reposts_count = retweet_sel[0].xpath(
'.//ul[@class="act s-fr"]/li/a[1]/text()'
).extract_first()
reposts_count = re.findall(r'\d+.*', reposts_count)
retweet['reposts_count'] = reposts_count[
0] if reposts_count else '0'
comments_count = retweet_sel[0].xpath(
'.//ul[@class="act s-fr"]/li[2]/a[1]/text()'
).extract_first()
comments_count = re.findall(r'\d+.*', comments_count)
retweet['comments_count'] = comments_count[
0] if comments_count else '0'
#Crawl Comments
if '万' in retweet['comments_count'] or int(
retweet['comments_count']) > 20:
#详细微博的url
comment_url = 'https://weibo.com/' + retweet[
'user_id'] + '/' + retweet['bid']
yield scrapy.Request(url=comment_url,
callback=self.parse_comments,
meta={
'weibo_id':
retweet['id'],
'weibo_bid':
retweet['bid'],
'weibo_user_id':
retweet['user_id'],
'keyword':
keyword
})
attitudes_count = retweet_sel[0].xpath(
'.//a[@action-type="feed_list_like"]/em/text()'
).extract_first()
retweet['attitudes_count'] = (attitudes_count
if attitudes_count else '0')
created_at = retweet_sel[0].xpath(
'.//p[@class="from"]/a[1]/text()').extract_first(
).replace(' ', '').replace('\n', '').split('前')[0]
retweet['created_at'] = util.standardize_date(created_at)
source = retweet_sel[0].xpath(
'.//p[@class="from"]/a[2]/text()').extract_first()
retweet['source'] = source if source else ''
retweet['pics'] = pics
retweet['video_url'] = video_url
retweet['retweet_id'] = ''
yield {'weibo': retweet, 'keyword': keyword}
weibo['retweet_id'] = retweet['id']
# print(weibo)
yield {'weibo': weibo, 'keyword': keyword}
def parse_comments(self, response):
MAX_COMM_NUM = 10000
WAIT_TIME = 1
weibo_id = response.meta.get('weibo_id')
weibo_bid = response.meta.get('weibo_bid')
weibo_user_id = response.meta.get('weibo_user_id')
keyword = response.meta.get('keyword')
'''chrome options'''
chrome_options = Options()
chrome_options.add_argument('--headless') # 使用无头谷歌浏览器模式
chrome_options.add_argument('--disable-gpu')
chrome_options.add_argument('--no-sandbox')
chrome_options.add_argument('log-level=2') #log级别设为error
#通知设置，否则浏览器会弹出通知框
prefs = {
'profile.default_content_setting_values': {
'notifications': 2
},
"profile.managed_default_content_settings.image": 2,
}
chrome_options.add_experimental_option('prefs', prefs)
driver = webdriver.Chrome(
chrome_options=chrome_options,
executable_path=
'E:\研三\weibo_crawl\chromedriver_win32/chromedriver.exe')
# 隐性等待1s
# driver.implicitly_wait(1)
#cookie
try:
driver.get(response.url)
time.sleep(WAIT_TIME)
if driver.current_url != response.url:
driver.delete_all_cookies()
for cookie in self.login_cookies:
driver.add_cookie(cookie)
driver.get(response.url)
time.sleep(WAIT_TIME)
except:
pass
#按热度排序
try:
WebDriverWait(driver, 3, 0.5).until(
EC.visibility_of_element_located((
By.XPATH,
'//div[@node-type="feed_cate"]/ul[@class="clearfix"]//a[@suda-uatrack="key=comment&value=hotcomm"]'
)))
hotcomm = driver.find_element_by_xpath(
'//div[@node-type="feed_cate"]/ul[@class="clearfix"]//a[@suda-uatrack="key=comment&value=hotcomm"]'
)
hotcomm.click()
time.sleep(WAIT_TIME)
except:
pass
try:
'''滚动到最后一条评论，循环滚动三次'''
js4 = "arguments[0].scrollIntoView();"
WebDriverWait(driver, 3, 0.5).until(
EC.visibility_of_element_located((
By.XPATH,
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]'
)))
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]')
driver.execute_script(js4, cur_last_comment)
time.sleep(WAIT_TIME)
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]')
driver.execute_script(js4, cur_last_comment)
time.sleep(WAIT_TIME)
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]')
driver.execute_script(js4, cur_last_comment)
time.sleep(WAIT_TIME)
#定位加载更多元素
WebDriverWait(driver, 3, 0.5).until(
EC.visibility_of_element_located((
By.XPATH,
'//div[@class="list_box"]/div[@class="list_ul"]/a[@action-type="click_more_comment"]'
)))
more_comment_ele = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/a[@action-type="click_more_comment"]'
)
#点击加载更多，最多点击1000次
for _ in range(1000):
more_comment_ele.click()
time.sleep(WAIT_TIME)
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]'
)
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]'
)
driver.execute_script(js4, cur_last_comment)
time.sleep(WAIT_TIME)
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]'
)
cur_last_comment = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div[last()]'
)
driver.execute_script(js4, cur_last_comment)
time.sleep(WAIT_TIME)
WebDriverWait(driver, 3, 0.5).until(
EC.visibility_of_element_located((
By.XPATH,
'//div[@class="list_box"]/div[@class="list_ul"]/a[@action-type="click_more_comment"]'
)))
more_comment_ele = driver.find_element_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/a[@action-type="click_more_comment"]'
)
except:
pass
#获取评论列表
try:
time.sleep(WAIT_TIME)
comments = driver.find_elements_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div')
comments = driver.find_elements_by_xpath(
'//div[@class="list_box"]/div[@class="list_ul"]/div')
total_num = len(comments)
except:
driver.quit()
return
#解析每条评论
crawled_cnt = 0
for comm_no in range(1, total_num + 1):
comment_item = CommentItem()
comment_item['weibo_id'] = weibo_id
comment_item['weibo_bid'] = weibo_bid
comment_item['weibo_user_id'] = weibo_user_id
xpath = '//div[@class="list_box"]/div[@class="list_ul"]/div[{0}]'.format(
comm_no)
ele = driver.find_element_by_xpath(xpath)
try:
comment_item['id'] = ele.get_attribute('comment_id')
except:
print('comment id test!')
try:
WB_text = ele.find_element_by_xpath(
'./div[@class="list_con"]/div[@class="WB_text"]')
except:
continue
comment_item['user_id'] = WB_text.find_element_by_xpath(
'./a[1]').get_attribute('href').split('/')[-1]
comment_item['screen_name'] = WB_text.find_element_by_xpath(
'./a[1]').text
#user_type
user_type = []
try:
title = WB_text.find_element_by_xpath(
'./a[@suda-data]/i').get_attribute('title')
user_type.append(title)
except:
pass
try:
type = WB_text.find_element_by_xpath(
'./a[@action-type]').get_attribute('title')
user_type.append(type)
except:
pass
if not user_type:
user_type = '其他'
else:
user_type = ';'.join(user_type)
comment_item['user_type'] = user_type
#attitudes_count
comment_item['attitudes_count'] = 0
try:
like_count = ele.find_element_by_xpath(
'.//ul[@class="clearfix"]/li//span[@node-type="like_status"]/em[last()]'
).text
if like_count != '赞':
comment_item['attitudes_count'] = like_count
except:
pass
#text
comment_item['text'] = WB_text.text
#created at
try:
created_at = ele.find_element_by_xpath(
'./div[@class="list_con"]/div[@class="WB_func clearfix"]/div[@class="WB_from S_txt2"]'
).text.replace(' ', '').replace('\n', '').split('前')[0]
comment_item['created_at'] = created_at
except:
comment_item['created_at'] = 'None'
yield {'comment': comment_item, 'keyword': keyword}
crawled_cnt += 1
if crawled_cnt > MAX_COMM_NUM:
break
driver.quit()
return
