【发布时间】：2016-12-15 22:06:51
【问题描述】：
我正在为 yellowpages.com 开发一个网络爬虫，总体上看起来运行良好。但是，在遍历长查询的分页时，requests.get(url) 将随机返回<Response [503]> 或<Response [404]>。偶尔，我会收到更严重的异常，例如：
requests.exceptions.ConnectionError: HTTPConnectionPool(host='www.yellowpages.com', port=80): 最大重试次数 超出网址： /search?search_terms=florists&geo_location_terms=FL&page=22（由 NewConnectionError(': 建立新连接失败: [WinError 10053] 已建立的连接被软件中止 在您的主机中'，））
使用 time.sleep() 似乎可以消除 503 错误，但 404 和异常仍然存在问题。
我正在尝试弄清楚如何“捕捉”各种响应，以便进行更改（等待、更改代理、更改用户代理）并重试和/或继续。伪代码如下：
If error/exception with request.get:
wait and/or change proxy and user agent
retry request.get
else:
pass
此时，我什至无法使用以下方法捕捉问题：
try:
r = requests.get(url)
except requests.exceptions.RequestException as e:
print (e)
import sys #only added here, because it's not part of my stable code below
sys.exit()
我从github 及以下开始的完整代码：
import requests
from bs4 import BeautifulSoup
import itertools
import csv
# Search criteria
search_terms = ["florists", "pharmacies"]
search_locations = ['CA', 'FL']
# Structure for Data
answer_list = []
csv_columns = ['Name', 'Phone Number', 'Street Address', 'City', 'State', 'Zip Code']
# Turns list of lists into csv file
def write_to_csv(csv_file, csv_columns, answer_list):
with open(csv_file, 'w') as csvfile:
writer = csv.writer(csvfile, lineterminator='\n')
writer.writerow(csv_columns)
writer.writerows(answer_list)
# Creates url from search criteria and current page
def url(search_term, location, page_number):
template = 'http://www.yellowpages.com/search?search_terms={search_term}&geo_location_terms={location}&page={page_number}'
return template.format(search_term=search_term, location=location, page_number=page_number)
# Finds all the contact information for a record
def find_contact_info(record):
holder_list = []
name = record.find(attrs={'class': 'business-name'})
holder_list.append(name.text if name is not None else "")
phone_number = record.find(attrs={'class': 'phones phone primary'})
holder_list.append(phone_number.text if phone_number is not None else "")
street_address = record.find(attrs={'class': 'street-address'})
holder_list.append(street_address.text if street_address is not None else "")
city = record.find(attrs={'class': 'locality'})
holder_list.append(city.text if city is not None else "")
state = record.find(attrs={'itemprop': 'addressRegion'})
holder_list.append(state.text if state is not None else "")
zip_code = record.find(attrs={'itemprop': 'postalCode'})
holder_list.append(zip_code.text if zip_code is not None else "")
return holder_list
# Main program
def main():
for search_term, search_location in itertools.product(search_terms, search_locations):
i = 0
while True:
i += 1
url = url(search_term, search_location, i)
r = requests.get(url)
soup = BeautifulSoup(r.text, "html.parser")
main = soup.find(attrs={'class': 'search-results organic'})
page_nav = soup.find(attrs={'class': 'pagination'})
records = main.find_all(attrs={'class': 'info'})
for record in records:
answer_list.append(find_contact_info(record))
if not page_nav.find(attrs={'class': 'next ajax-page'}):
csv_file = "YP_" + search_term + "_" + search_location + ".csv"
write_to_csv(csv_file, csv_columns, answer_list) # output data to csv file
break
if __name__ == '__main__':
main()
提前感谢您抽出宝贵时间阅读这篇长篇文章/回复:)
【问题讨论】：
标签： python exception-handling web-scraping python-requests yellow-pages