def open_spider(self, spider):
#获取数据库中微博内容最新时间戳
if spider.name in ('keyweibocontent', 'weibocontent_userinfo',
'weibocontent_danger',
'weibocontent_userinfo_intime'):
db = OracleStore()
conn = db.get_connection()
sql = "select * from t_user_weibocontent where userID = '%s' order by publishTimeStamp desc" % str(
spider.uid)
cursor = db.select_operation(conn, sql)
count = cursor.fetchone()
if not count: #count为None，即数据库中没有该用户微博及时间戳的数据
WeibospiderPipeline.weibocontent_timestamp = None
else:
WeibospiderPipeline.weibocontent_timestamp = count[
6] #获取数据库中最新的时间戳（publishTimeStamp）字段
db.close_connection(conn, cursor)
if spider.name == 'keyuser':
db = OracleStore()
conn = db.get_connection()
sql = "select * from t_user_keyword where keyword = '%s' order by publishTimeStamp desc" % str(
spider.keyword)
cursor = db.select_operation(conn, sql)
count = cursor.fetchone()
if not count: #count为None，即数据库中没有该关键词搜索结果及时间戳的数据
WeibospiderPipeline.keyword_timestamp = None
else:
WeibospiderPipeline.keyword_timestamp = count[
4] #获取数据库中最新的时间戳（publishTimeStamp）字段
db.close_connection(conn, cursor)
def start_getweiboinfo(self,response):
db = OracleStore();conn = db.get_connection()
sql1 = '''select * from t_user_info'''
cursor1 = db.select_operation(conn,sql1)
sql2 = '''select count(*) from t_user_info'''
cursor2 = db.select_operation(conn,sql2)
count = cursor2.fetchone()
if count[0]:
for i in range(count[0]):
for result in cursor1.fetchmany(1):
mainpageurl = 'http://weibo.com/u/'+str(result[0])+'?from=otherprofile&wvr=3.6&loc=tagweibo'
GetWeibopage.data['uid'] = result[0]
getweibopage = GetWeibopage()
for page in range(WeiboSpider.page_num):
GetWeibopage.data['page'] = page+1
firstloadurl = mainpageurl + getweibopage.get_firstloadurl()
yield Request(url=firstloadurl,meta={'cookiejar':response.meta['cookiejar'],'uid':result[0]},callback=self.parse_load)
secondloadurl = mainpageurl + getweibopage.get_secondloadurl()
yield Request(url=secondloadurl,meta={'cookiejar':response.meta['cookiejar'],'uid':result[0]},callback=self.parse_load)
thirdloadurl = mainpageurl + getweibopage.get_thirdloadurl()
yield Request(url=thirdloadurl,meta={'cookiejar':response.meta['cookiejar'],'uid':result[0]},callback=self.parse_load)
else:
yield None
db.close_connection(conn,cursor1,cursor2)
def parse_load(self, response):
user_info = userinfo.WeiboSpider()
request_url = response.request.url
p = re.compile('&pre_page=(\d).*&page=(\d)') #用于判断是第一页的第一次加载
match = p.search(request_url)
if int(match.group(1)) == 0 and int(match.group(2)) == 1: #进行用户信息的获取
db = OracleStore()
conn = db.get_connection()
sql = "select count(*) from t_user_info where userID='%s'" % self.uid
cursor = db.select_operation(conn, sql)
count = cursor.fetchone()
if not count[0]: #若没有爬取过该uid用户,则爬取用户基本信息
analyzer = Analyzer()
total_pq = analyzer.get_html(
response.body, 'script:contains("PCD_person_info")')
user_property = analyzer.get_userproperty(total_pq)
if user_property == 'icon_verify_co_v': #该账号为公众账号
public_userinfo_url = analyzer.get_public_userinfohref(
total_pq)
#yield Request(url=public_userinfo_url,meta={'cookiejar':response.meta['cookiejar'],'uid':response.meta['uid'],'user_property':user_property},callback=self.parse_public_userinfo) 暂时不处理公众账号，需要数据库设置外键
else:
userinfo_url = analyzer.get_userinfohref(total_pq)
yield Request(url=userinfo_url,
meta={
'cookiejar': response.meta['cookiejar'],
'uid': response.meta['uid'],
'user_property': user_property
},
callback=self.parse_userinfo)
db.close_connection(conn, cursor)
item = WeibospiderItem() #获取用户微博信息及@用户信息
analyzer = Analyzer()
friendcircle = FriendCircle()
total_pq = analyzer.get_mainhtml(response.body)
item['uid'] = response.meta['uid']
item['content'] = analyzer.get_content(total_pq)
item['time'], item['timestamp'] = analyzer.get_time(total_pq)
atuser_info, item['repost_user'] = analyzer.get_atuser_repostuser(
total_pq)
atuser_list = friendcircle.atuser_parser(atuser_info)
item['atuser_nickname_list'] = atuser_list
yield item
for atuser_inlist in atuser_list:
if atuser_inlist != []:
for atuser in atuser_inlist:
uid_url = "http://s.weibo.com/user/" + quote(
quote(str(atuser))) + "&Refer=SUer_box"
yield Request(url=uid_url,
meta={
'cookiejar': response.meta['cookiejar'],
'uid': self.uid,
'atuser_nickname': atuser
},
callback=self.parse_atuser_uid)
else:
continue
def start_getweiboinfo(self, response):
db = OracleStore()
conn = db.get_connection()
sql1 = '''select * from t_user_info'''
cursor1 = db.select_operation(conn, sql1)
sql2 = '''select count(*) from t_user_info'''
cursor2 = db.select_operation(conn, sql2)
count = cursor2.fetchone()
if count[0]:
for i in range(count[0]):
for result in cursor1.fetchmany(1):
mainpageurl = 'http://weibo.com/u/' + str(
result[0]) + '?from=otherprofile&wvr=3.6&loc=tagweibo'
GetWeibopage.data['uid'] = result[0]
getweibopage = GetWeibopage()
for page in range(WeiboSpider.page_num):
GetWeibopage.data['page'] = page + 1
firstloadurl = mainpageurl + getweibopage.get_firstloadurl(
)
yield Request(url=firstloadurl,
meta={
'cookiejar':
response.meta['cookiejar'],
'uid': result[0]
},
callback=self.parse_load)
secondloadurl = mainpageurl + getweibopage.get_secondloadurl(
)
yield Request(url=secondloadurl,
meta={
'cookiejar':
response.meta['cookiejar'],
'uid': result[0]
},
callback=self.parse_load)
thirdloadurl = mainpageurl + getweibopage.get_thirdloadurl(
)
yield Request(url=thirdloadurl,
meta={
'cookiejar':
response.meta['cookiejar'],
'uid': result[0]
},
callback=self.parse_load)
else:
yield None
db.close_connection(conn, cursor1, cursor2)
def parse_load(self,response):
user_info = userinfo.WeiboSpider()
request_url = response.request.url
p=re.compile('&pre_page=(\d).*&page=(\d)') #用于判断是第一页的第一次加载
match = p.search(request_url)
if int(match.group(1)) == 0 and int(match.group(2)) == 1: #进行用户信息的获取
db = OracleStore();conn = db.get_connection()
sql = "select count(*) from t_user_info where userID='%s'" % self.uid
cursor = db.select_operation(conn,sql);count = cursor.fetchone()
if not count[0]: #若没有爬取过该uid用户,则爬取用户基本信息
analyzer = Analyzer()
total_pq = analyzer.get_html(response.body,'script:contains("PCD_person_info")')
user_property = analyzer.get_userproperty(total_pq)
if user_property == 'icon_verify_co_v': #该账号为公众账号
public_userinfo_url = analyzer.get_public_userinfohref(total_pq)
#yield Request(url=public_userinfo_url,meta={'cookiejar':response.meta['cookiejar'],'uid':response.meta['uid'],'user_property':user_property},callback=self.parse_public_userinfo) 暂时不处理公众账号，需要数据库设置外键
else:
userinfo_url = analyzer.get_userinfohref(total_pq)
yield Request(url=userinfo_url,meta={'cookiejar':response.meta['cookiejar'],'uid':response.meta['uid'],'user_property':user_property},callback=self.parse_userinfo)
db.close_connection(conn,cursor)
item = WeibospiderItem() #获取用户微博信息及@用户信息
analyzer = Analyzer()
friendcircle = FriendCircle()
total_pq = analyzer.get_mainhtml(response.body)
item['uid'] = response.meta['uid']
item['content'] = analyzer.get_content(total_pq)
item['time'],item['timestamp'] = analyzer.get_time(total_pq)
atuser_info,item['repost_user'] = analyzer.get_atuser_repostuser(total_pq)
atuser_list = friendcircle.atuser_parser(atuser_info)
item['atuser_nickname_list'] = atuser_list
yield item
for atuser_inlist in atuser_list:
if atuser_inlist != []:
for atuser in atuser_inlist:
uid_url = "http://s.weibo.com/user/"+quote(quote(str(atuser)))+"&Refer=SUer_box"
yield Request(url=uid_url,meta={'cookiejar':response.meta['cookiejar'],'uid':self.uid,'atuser_nickname':atuser},callback=self.parse_atuser_uid)
else:
continue
def open_spider(self,spider):
#获取数据库中微博内容最新时间戳
if spider.name in ('keyweibocontent','weibocontent_userinfo','weibocontent_danger','weibocontent_userinfo_intime'):
db=OracleStore();conn = db.get_connection()
sql = "select * from t_user_weibocontent where userID = '%s' order by publishTimeStamp desc" % str(spider.uid)
cursor = db.select_operation(conn,sql)
count = cursor.fetchone()
if not count: #count为None，即数据库中没有该用户微博及时间戳的数据
WeibospiderPipeline.weibocontent_timestamp = None
else:
WeibospiderPipeline.weibocontent_timestamp = count[6] #获取数据库中最新的时间戳（publishTimeStamp）字段
db.close_connection(conn,cursor)
if spider.name == 'keyuser':
db=OracleStore();conn = db.get_connection()
sql = "select * from t_user_keyword where keyword = '%s' order by publishTimeStamp desc" % str(spider.keyword)
cursor = db.select_operation(conn,sql)
count = cursor.fetchone()
if not count: #count为None，即数据库中没有该关键词搜索结果及时间戳的数据
WeibospiderPipeline.keyword_timestamp = None
else:
WeibospiderPipeline.keyword_timestamp = count[4] #获取数据库中最新的时间戳（publishTimeStamp）字段
db.close_connection(conn,cursor)
def closed(self, reason):
db = OracleStore()
conn = db.get_connection()
db.close_connection(conn)
print "--------closed-------"
def closed(self, reason):
db = OracleStore()
conn = db.get_connection()
db.close_connection(conn)
print '--------closed-------'