def parse_page(self, response):
""" 抓取微博数据 """
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
id = tweet.xpath('@id').extract_first() # 微博ID
content = tweet.xpath(
'div/span[@class="ctt"]/text()').extract_first() # 微博内容
ctime = tweet.xpath(
'div/span[@class="ct"]/text()').extract() #微博创建时间
cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
allinks = re.findall(u"<a.*?href=.*?[\d+].*?<\/a>",
tweet.extract(), re.I | re.S | re.M)
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]',
tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]',
tweet.extract()) # 评论数
comment_link = ""
for link in allinks:
like_link_list = re.findall(u'"(http.*)">\u8d5e.*', link)
if len(like_link_list) > 0:
like_link = like_link_list[0]
transfer_link_list = re.findall(u'"(http.*)">\u8f6c\u53d1.*',
link)
if len(transfer_link_list) > 0:
transfer_link = transfer_link_list[0]
comment_link_list = re.findall(
ur'"(http.*?)" class="cc".*>\u8bc4\u8bba\[\d+\]</a>', link)
if len(comment_link_list) > 0:
comment_link = comment_link_list[0]
tweetsItems = TweetsItem()
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first(
) # 求时间和使用工具（手机或平台）
tweetsItems["ID"] = "cctv"
if content:
tweetsItems["Content"] = content.strip(
u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
if ctime:
tweetsItems["Time"] = ctime
if cooridinates:
cooridinates = re.findall('center=([\d|.|,]+)', cooridinates)
if cooridinates:
tweetsItems["Co_oridinates"] = cooridinates[0]
if like:
tweetsItems["Like"] = int(like[0])
if transfer:
tweetsItems["Transfer"] = int(transfer[0])
if comment:
tweetsItems["Comment"] = int(comment[0])
if comment_link != "":
yield Request(url=comment_link,
meta={
'tweetsItems': tweetsItems,
'comment_link': comment_link
},
callback=self.parse_fans_name)
def parse2(self, response):
""" 抓取微博数据 """
if response.body == "":
req = response.request
req.meta["change_proxy"] = True
yield req
else:
# logger.info("got page: %s" % response.body)
selector = Selector(response)
urllist = selector.xpath('//a[@href]/@href').extract()
for text00 in urllist:
reposturl = re.findall(r'^http://weibo.cn/repost/',text00)
if reposturl:
if text00 not in self.finishurllist:
self.starturllist.add(text00)
# 爬取ID
text7 = selector.xpath('body/div[@class="u"]/div[@class="tip2"]').extract_first()
if text7:
ID = re.findall('uid=(\d+)', text7)[0]
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
tweetsItems = TweetsItem()
id = tweet.xpath('@id').extract_first() # 微博ID
content = tweet.xpath('div/span[@class="ctt"]/text()').extract_first() # 微博内容
cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]', tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]', tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first() # 求时间和使用工具（手机或平台）
tweetsItems["ID"] = ID
tweetsItems["_id"] = ID + "-" + id
tweetsItems["tweetID"] = id
if content:
tweetsItems["Content"] = content.strip(u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
if cooridinates:
cooridinates = re.findall('center=([\d|.|,]+)', cooridinates)
if cooridinates:
tweetsItems["Co_oridinates"] = cooridinates[0]
if like:
tweetsItems["Like"] = int(like[0])
if transfer:
tweetsItems["Transfer"] = int(transfer[0])
if comment:
tweetsItems["Comment"] = int(comment[0])
if others:
others = others.split(u"\u6765\u81ea")
tweetsItems["PubTime"] = others[0]
if len(others) == 2:
tweetsItems["Tools"] = others[1]
yield tweetsItems
def parse2(self, response):
""" 抓取微博数据 """
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
tweetsItems = TweetsItem()
id = tweet.xpath('@id').extract_first() # 微博ID
content = tweet.xpath(
'div/span[@class="ctt"]/text()').extract_first() # 微博内容
cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]',
tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]',
tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first(
) # 求时间和使用工具（手机或平台）
if others:
others = others.split(u"\u6765\u81ea")
hasMonth = re.findall(u'\u6708', others[0])
if hasMonth:
temp = others[0].split(u'\u6708')
month = temp[0]
day = temp[1].split(u'\u65e5')[0]
if int(month) == int(self.current_month) and (
int(self.current_day) - int(day)) <= 1:
tweetsItems["PubTime"] = others[0]
if len(others) == 2:
tweetsItems["Tools"] = others[1]
else:
continue
else:
continue
if content:
choujiang = re.findall(u'\u62bd', content)
if not choujiang:
continue
tweetsItems["Content"] = content.strip(
u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
tweetsItems["ID"] = response.meta["ID"]
tweetsItems["_id"] = id
if cooridinates:
cooridinates = re.findall('center=([\d|.|,]+)', cooridinates)
if cooridinates:
tweetsItems["Co_oridinates"] = cooridinates[0]
if like:
tweetsItems["Like"] = int(like[0])
if transfer:
tweetsItems["Transfer"] = int(transfer[0])
if comment:
tweetsItems["Comment"] = int(comment[0])
yield tweetsItems
def parse2(self, response):
""" 抓取微博数据 """
# print("========33333333=======")
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
# print(tweets)
for tweet in tweets:
tweetsItems = TweetsItem()
id = tweet.xpath('@id').extract_first() # 微博ID
content = tweet.xpath(
'div/span[@class="ctt"]/text()').extract_first() # 微博内容
cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]',
tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]',
tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first(
) # 求时间和使用工具（手机或平台）
tweetsItems["ID"] = response.meta["ID"]
tweetsItems["_id"] = response.meta["ID"] + "-" + id
if content:
tweetsItems["Content"] = content.strip(
u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
if cooridinates:
cooridinates = re.findall('center=([\d|.|,]+)', cooridinates)
if cooridinates:
tweetsItems["Co_oridinates"] = cooridinates[0]
if like:
tweetsItems["Like"] = int(like[0])
if transfer:
tweetsItems["Transfer"] = int(transfer[0])
if comment:
tweetsItems["Comment"] = int(comment[0])
if others:
others = others.split(u"\u6765\u81ea")
tweetsItems["PubTime"] = others[0]
if len(others) == 2:
tweetsItems["Tools"] = others[1]
# print(tweetsItems)
yield tweetsItems
url_next = selector.xpath(
u'body/div[@class="pa" and @id="pagelist"]/form/div/a[text()="\u4e0b\u9875"]/@href'
).extract()
if url_next:
yield Request(url=self.host + url_next[0],
meta={"ID": response.meta["ID"]},
callback=self.parse2)
def parse_weibo1(self, response):
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
tweetsItem = TweetsItem()
weibo_id = tweet.xpath('@id').extract_first()[2:] # 微博ID
cmts = tweet.xpath('div/span[@class="cmt"]').extract()
if len(tweet.xpath('div/span[@class="cmt"]').extract()) > 2:
r = tweet.xpath(u'div/span[text() = "转发理由:"]')
content = r.xpath('./../text()').extract_first()
like = re.findall(u'\u8d5e\[(\d+)\]', ','.join(r.xpath(u'./../a/text()').extract())) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]', ','.join(r.xpath(u'./../a/text()').extract())) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]', ','.join(r.xpath(u'./../a/text()').extract())) # 评论数
tweetsItem['Type'] = REPOST
coordinates = None
else:
content = tweet.xpath('div/span[@class="ctt"]/text()').extract_first() # 微博内容
coordinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
tweetsItem['Type'] = ORIGINAL
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]', tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]', tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first() # 求时间和使用工具（手机或平台）
tweetsItem["ID"] = response.meta["ID"]
tweetsItem["_id"] = response.meta["ID"] + "-" + weibo_id
if content:
tweetsItem["Content"] = content.strip(u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
if coordinates:
coordinates = re.findall('center=([\d|.|,]+)', coordinates)
if coordinates:
tweetsItem["Coordinates"] = coordinates[0]
if like:
tweetsItem["Like"] = int(like[0])
if transfer:
tweetsItem["Transfer"] = int(transfer[0])
if comment:
tweetsItem["Comment"] = int(comment[0])
if others:
others = others.split(u"\u6765\u81ea")
tweetsItem["PubTime"] = others[0]
if len(others) == 2:
tweetsItem["Tools"] = others[1]
# yield Request(url=comment_url, meta={"weiboId": weibo_id, "current_page": 1}, callback=self.parse_comment)
yield tweetsItem
def parse2(self, response):
""" 抓取微博数据 """
selector = Selector(response)
# g = selector.xpath('//input[@name="mp"]/@value').extract_first()
# mp = int(g)
# b = response.url.split("=")
# k = b[1]
# j = int(k)
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
tweetsItems = TweetsItem()
id = tweet.xpath('@id').extract_first() # 微博ID
ctt = tweet.xpath('div/span[@class="ctt"]')
content = ctt.xpath('string(.)').extract()[0]
# content = tweet.xpath('div/span[@class="ctt"]/text()').extract_first() # 微博内容
# cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]', tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]', tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first() # 求时间和使用工具（手机或平台）
tweetsItems["ID"] = response.meta["ID"]
tweetsItems["_id"] = response.meta["ID"] + "-" + id
if content:
tweetsItems["Content"] = content
# tweetsItems["Content"] = content.strip(u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
# if cooridinates:
# cooridinates = re.findall('center=([\d|.|,]+)', cooridinates)
# if cooridinates:
# tweetsItems["Co_oridinates"] = cooridinates[0]
if like:
tweetsItems["Like"] = int(like[0])
if transfer:
tweetsItems["Transfer"] = int(transfer[0])
if comment:
tweetsItems["Comment"] = int(comment[0])
if others:
others = others.split(u"\u6765\u81ea")
tweetsItems["PubTime"] = others[0]
# if len(others) == 2:
# tweetsItems["Tools"] = others[1]
yield tweetsItems
def parse2(self, response):
""" 抓取微博数据 """
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
tweetsItems = TweetsItem()
id = tweet.xpath('@id').extract_first() # 微博ID
content = tweet.xpath(
'div/span[@class="ctt"]/text()').extract_first() # 微博内容
cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first(
) # 求时间和使用工具（手机或平台）
tweetsItems["ID"] = response.meta["ID"]
tweetsItems["_id"] = response.meta["ID"] + "-" + id
if content:
tweetsItems["Content"] = content.strip(
u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
else:
tweetsItems["Content"] = u""
if others:
others = others.split(u"\u6765\u81ea")
tweetsItems["PubTime"] = others[0]
if len(others) == 2:
tweetsItems["Tools"] = others[1]
else:
tweetsItems["Tools"] = u""
yield tweetsItems
url_next = selector.xpath(
u'body/div[@class="pa" and @id="pagelist"]/form/div/a[text()="\u4e0b\u9875"]/@href'
).extract()
if url_next:
yield Request(url=self.host + url_next[0],
meta={"ID": response.meta["ID"]},
callback=self.parse2)
print(self.host + url_next[0])
def parse_weibo(self, response):
""" 抓取微博数据 """
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
for tweet in tweets:
tweetsItem = TweetsItem()
weibo_id = tweet.xpath('@id').extract_first()[2:] # 微博ID
cmts = tweet.xpath('div/span[@class="cmt"]').extract()
if len(tweet.xpath('div/span[@class="cmt"]').extract()) > 2:
content = tweet.xpath(
u'div/span[text() = "转发理由:"]/../text()').extract_first()
tweetsItem['Type'] = REPOST
coordinates = None
else:
content = tweet.xpath(
'div/span[@class="ctt"]/text()').extract_first() # 微博内容
coordinates = tweet.xpath(
'div/a/@href').extract_first() # 定位坐标
tweetsItem['Type'] = ORIGINAL
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]',
tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]',
tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first(
) # 求时间和使用工具（手机或平台）
tweetsItem["ID"] = response.meta["ID"]
comment_url = self.comment_pattern % (weibo_id, 1)
tweetsItem["_id"] = response.meta["ID"] + "-" + weibo_id
if content:
tweetsItem["Content"] = content.strip(
u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
if coordinates:
coordinates = re.findall('center=([\d|.|,]+)', coordinates)
if coordinates:
tweetsItem["Coordinates"] = coordinates[0]
if like:
tweetsItem["Like"] = int(like[0])
if transfer:
tweetsItem["Transfer"] = int(transfer[0])
if comment:
tweetsItem["Comment"] = int(comment[0])
if others:
others = others.split(u"\u6765\u81ea")
tweetsItem["PubTime"] = others[0]
if len(others) == 2:
tweetsItem["Tools"] = others[1]
yield Request(url=comment_url,
meta={
"weiboId": weibo_id,
"current_page": 1
},
callback=self.parse_comment)
yield tweetsItem
next_page = response.meta['current_page'] + 1
if 'max_page' not in response.meta:
print selector.xpath('body/div[@id="pagelist"]')
max_page = selector.xpath(
'body/div[@id="pagelist"]/form/div/input[@name="mp"]/@value'
).extract_first()
if max_page:
response.meta['max_page'] = int(max_page)
else:
return
print response.meta['max_page']
if next_page <= response.meta['max_page']:
response.meta['current_page'] = next_page
yield Request(url=self.host + '/' + response.meta['ID'] +
'?page=' + str(next_page),
meta=response.meta,
callback=self.parse_weibo)
def parse2(self, response):
""" 抓取微博数据 """
selector = Selector(response)
tweets = selector.xpath('body/div[@class="c" and @id]')
ctime = datetime.now()
crawl = True
# lastCrawlTime = self.get_lastWeiboDate(response.meta["ID"])
for tweet in tweets:
tweetsItems = TweetsItem()
id = tweet.xpath('@id').extract_first() # 微博ID
content = "" # 微博内容
for c in tweet.xpath(
'div/span[@class="ctt"]/text() | div/span[@class="ctt"]/a/text()'
'| div/a/text() | div/text()'):
content += c.extract() # 微博内容
content = content.split("\xa0")[0].replace("#",
"").replace("全文", "")
cooridinates = tweet.xpath('div/a/@href').extract_first() # 定位坐标
like = re.findall(u'\u8d5e\[(\d+)\]', tweet.extract()) # 点赞数
transfer = re.findall(u'\u8f6c\u53d1\[(\d+)\]',
tweet.extract()) # 转载数
comment = re.findall(u'\u8bc4\u8bba\[(\d+)\]',
tweet.extract()) # 评论数
others = tweet.xpath('div/span[@class="ct"]/text()').extract_first(
) # 求时间和使用工具（手机或平台）
tweetsItems["ID"] = response.meta["ID"]
tweetsItems["_id"] = response.meta["ID"] + "-" + id
if content:
tweetsItems["Content"] = content.strip(
u"[\u4f4d\u7f6e]") # 去掉最后的"[位置]"
if cooridinates:
cooridinates = re.findall('center=([\d|.|,]+)', cooridinates)
if cooridinates:
tweetsItems["Co_oridinates"] = cooridinates[0]
if like:
tweetsItems["Like"] = int(like[0])
if transfer:
tweetsItems["Transfer"] = int(transfer[0])
if comment:
tweetsItems["Comment"] = int(comment[0])
if others:
others = others.split(u"\u6765\u81ea")
others[0] = others[0].replace(u"\xa0", '')
if "分钟前" in others[0]:
tweetsItems["PubTime"] = (
datetime.now() -
timedelta(minutes=int(others[0].split("分钟前")[0]))
).strftime("%Y-%m-%d %H:%M:%S")
elif "今天" in others[0]:
tweetsItems["PubTime"] = others[0].replace(
"今天",
datetime.now().strftime("%Y-%m-%d")) + ":00"
elif "月" in others[0]:
tweetsItems["PubTime"] = (str(datetime.now().year) + "-" +
others[0] + ":00").replace(
"月", "-").replace("日", '')
else:
tweetsItems["PubTime"] = others[0]
# if len(others) == 2:
# tweetsItems["Tools"] = others[1]
# 提取hashtag
tweetsItems["Label"] = tweet.xpath(
'div/span[@class="ctt"]/a/text()').extract_first() # hashtag
ctime = datetime.strptime(tweetsItems["PubTime"],
"%Y-%m-%d %H:%M:%S")
if self.yetNotCrawled(tweetsItems["_id"]) == False:
crawl = False
break
yield tweetsItems
url_next = selector.xpath(
u'body/div[@class="pa" and @id="pagelist"]/form/div/a[text()="\u4e0b\u9875"]/@href'
).extract()
if url_next and crawl and self.tweetTimeDelta(ctime) <= 30:
yield Request(url=self.host + url_next[0],
meta={"ID": response.meta["ID"]},
callback=self.parse2)
