Question

我尝试组织最多10次并发下载的池。该函数应该下载基本URL，然后解析该页面上的所有URL并下载每个URL，但是并发下载的总数不应超过10个。

from lxml import etree 
import gevent
from gevent import monkey, pool
import requests

monkey.patch_all()
urls = [
    'http://www.google.com', 
    'http://www.yandex.ru', 
    'http://www.python.org', 
    'http://stackoverflow.com',
    # ... another 100 urls
    ]

LINKS_ON_PAGE=[]
POOL = pool.Pool(10)

def parse_urls(page):
    html = etree.HTML(page)
    if html:
        links = [link for link in html.xpath("//a/@href") if 'http' in link]
    # Download each url that appears in the main URL
    for link in links:
        data = requests.get(link)
        LINKS_ON_PAGE.append('%s: %s bytes: %r' % (link, len(data.content), data.status_code))

def get_base_urls(url):
    # Download the main URL
    data = requests.get(url)
    parse_urls(data.content)

如何组织它以并发方式，但保持所有Web请求的一般全局池限制？

Answer 1

gevent.pool将限制并发的greenlets，而不是连接。

您应该将session与HTTPAdapter

一起使用

connection_limit = 10
adapter = requests.adapters.HTTPAdapter(pool_connections=connection_limit, 
                                        pool_maxsize=connection_limit)
session = requests.session()
session.mount('http://', adapter)
session.get('some url')
# or do your work with gevent
from gevent.pool import Pool
# it should bigger than connection limit if the time of processing data 
# is longer than downings, 
# to give a change run processing.
pool_size = 15 
pool = Pool(pool_size)
for url in urls:
    pool.spawn(session.get, url)

Answer 2

我认为以下内容应该能满足您的需求。我在我的例子中使用的是BeautifulSoup，而不是链接条带化的东西。

from bs4 import BeautifulSoup
import requests
import gevent
from gevent import monkey, pool
monkey.patch_all()

jobs = []
links = []
p = pool.Pool(10)

urls = [
    'http://www.google.com', 
    # ... another 100 urls
]

def get_links(url):
    r = requests.get(url)
    if r.status_code == 200:
        soup = BeautifulSoup(r.text)
        links + soup.find_all('a')

for url in urls:
    jobs.append(p.spawn(get_links, url))
gevent.joinall(jobs)

Answer 3

您应该使用gevent.queue以正确的方式执行此操作。

同样this(eventlet examples)将有助于您理解基本理念。

Gevent解决方案类似于eventlet。

请记住，有一些地方可以存储访问过的网址，以免出现循环，因此您不会出现内存错误，需要引入一些限制。

具有嵌套Web请求的Gevent池

3 个答案: