|
今天爬取的是一个航空公司网站的航班信息–飞常准 我们需要遍历抓取每一个航班里的所有信息
  这里需要注意 有些列出来的航班可能不存在信息 需要进行一次判断 另外 访问过快会导致被该网站封号，解封会出现验证码，可以自己搭建免费ip代理池或使用收费代理 代码如下：
import requests
import json
import pytesseract
import re
import time
import random
from urllib.parse import urlencode
from lxml import etree
from PIL import Image
class FeiChangZhun(object):
def __init__(self):
self.base_url = 'http://www.variflight.com' # 飞常准官网
def get_list(self):
r = requests.get('http://www.variflight.com/sitemap.html?AE71649A58c77=') # 飞常准航班列表
selector = etree.HTML(r.text)
url_list = selector.xpath('//*[@class="list"]/a/@href')[1:]
# 解析出每一个航班下面的Url链接
new_list = []
for i in url_list:
new_list.append(self.base_url + i)
# 测试第一个航班数据
# mytest = new_list[9]
return new_list
def parse_html(self):
mytest = self.get_list()
proxy = self.get_ip()
for test in mytest:
try:
proxies = {"http": random.choice(proxy)}
r = requests.get(test, proxies=proxies)
selector = etree.HTML(r.text)
b = selector.xpath('//*[@class="searchlist_innerli"]')
if b:
name = selector.xpath('//*[@class="tit"]/h1/@title')[0]
log = "{0}航班存在信息：".format(name)
print(log)
# time.sleep(5)
r = requests.Session()
r.proxies = proxies
resp = r.get(test)
selector = etree.HTML(resp.text)
mylist = selector.xpath('//*[@id="list"]/li')
# print(mylist)
for selector in mylist:
a = selector.xpath('div[@class="li_com"]/span[1]/b/a//text()') # 航班信息
a = a[0] + '' + a[1]
b = selector.xpath('div[@class="li_com"]/span[2]/@dplan') # 计划起飞
c = selector.xpath('div[@class="li_com"]/span[3]/img/@src') # 实际起飞
if c:
url = self.base_url + c[0]
resp = r.get(url)
filename = './pictures' + re.search(r's=(.*?)==', url).group(0) + '.png'
with open(filename, 'wb') as f:
f.write(resp.content)
c = pytesseract.image_to_string(Image.open(filename))
if len(c) < 5: # 若识别不出‘:’或者‘.’ 进行拼接
c = c[:2] + ':' + c[2:]
else:
# c = selector.xpath('div[@class="li_com"]/span[3]/text()') # 实际起飞
# c = re.sub(r"[\s+\.\!\/_,$%^*(+\"\')]+|[+——?【】？~@#￥%……&*]+|\\n+|\\r+|(\\xa0)+|(\\u3000)+|\\t", "", str(c[0]))
c = '暂无信息'
d = selector.xpath('div[@class="li_com"]/span[4]/text()') # 出发地
e = selector.xpath('div[@class="li_com"]/span[5]/@aplan') # 计划到达
f = selector.xpath('div[@class="li_com"]/span[6]/text()') # 实际到达
f = re.sub(r"[\s+\.\!\/_,$%^*(+\"\')]+|[+?【】？~@#￥%……&*]+|\\n+|\\r+|(\\xa0)+|(\\u3000)+|\\t", "",
str(f[0]))
if f:
f = '暂无信息'
else:
f = selector.xpath('div[@class="li_com"]/span[6]/img/@src') # 实际到达
url = self.base_url + f[0]
resp = r.get(url)
filename = './pictures' + re.search(r's=(.*?)==', url).group(0) + '.png'
with open(filename, 'wb') as f:
f.write(resp.content)
f = pytesseract.image_to_string(Image.open(filename))
if len(f) < 5:
f = f[:2] + ':' + f[2:]
g = selector.xpath('div[@class="li_com"]/span[7]/text()') # 到达地
h = selector.xpath('div[@class="li_com"]/span[8]/img/@src') # 准点率
i = selector.xpath('div[@class="li_com"]/span[9]/text()') # 状态
j = selector.xpath('div[@class="li_com"]/span[10]/a[1]/text()') # 订制
h = self.base_url + h[0] # 准点率
filename = './pictures' + re.search(r's=(.*?)=', h).group(0) + '.png'
q = r.get(h)
with open(filename, 'wb') as t:
t.write(q.content)
q = pytesseract.image_to_string(Image.open(filename))
if len(q) < 5:
q = q[:2] + ':' + q[2:]
mydict = {
"title": a, # 航班信息
"start_time": b[0], # 计划起飞
"actual_start_time": c, # 实际起飞
"start_place": d[0], # 出发地
"arrive_time": e[0], # 计划到达
"actual_arrive_time": f, # 实际到达
"arrive_place": g[0], # 到达地
"on-time": q, # 准点率
"satus": i[0], # 状态
"order": j[0] # 定制
}
print(mydict)
else:
name = selector.xpath('//*[@id="byNumInput"]/@value')[0]
log = "{0}航班不存在信息".format(name)
print(log)
except:
continue
# ip代理池 根据自身情况选择免费或付费
def get_ip(self):
r = requests.get(
'http://piping.mogumiao.com/proxy/api/get_ip_bs?appKey=63a4113d5b2842869c9d2774b960cc3f&count=10&expiryDate=0&format=1&newLine=2')
mylist = r.text
myjson = json.loads(mylist)
mylist = myjson.get('msg')
pro = []
for i in mylist:
ip = "{0}:{1}".format(i.get("ip"), i.get("port"))
try:
a = requests.get('https://www.baidu.com/', proxies={'https': ip}, timeout=1)
if a.status_code == 200:
pro.append(ip)
except:
continue
return pro
def main():
Fei = FeiChangZhun()
Fei.parse_html()
if __name__ == '__main__':
main()
效果图如下： 
由于航班信息是实时性较强的，所以使用原生的requests来写爬虫会比使用框架更好更方便。
|