如何使用 webdriver 並行化 Python 循環？

Question

在下面的代碼中，我嘗試從不同的 url 頁面（總共 25 個）導入所有賠率。 我目前使用一個簡單的循環，但它需要太多時間。 如何並行化此代碼以減少執行時間。

這是代碼：

#!/usr/bin/python3
# -*- coding: utf­-8 ­-*-

from selenium import webdriver
import statistics as stat
import numpy as np

driver = webdriver.Firefox()
url = 'https://www.coteur.com/cotes-foot.php'
driver.get(url)

#Store url associated with the soccer games
url_links = []
for i in driver.find_elements_by_xpath('//a[contains(@href, "match/cotes-")]'):
    url_links.append(i.get_attribute('href'))

driver.close()
print(len(url_links), '\n')

for l in range(len(url_links)):
    driver = webdriver.Firefox()
    driver.get(url_links[l])

    #Store odds into table
    odds = []
    header = []
    for i in driver.find_elements_by_xpath('//button[contains(@class, "btn btn-default btn-xs btncote")]'):
        odds.append(i.text)

    for i in driver.find_elements_by_xpath('//th[contains(@width, "20%")]'):
        header.append(i.text)

    rows = int(len(odds)/3) 
    columns = 3
    odds = [float(i) for i in odds]
    odds = np.array(odds)
    header = np.array(header)
    odds = odds.reshape(rows, columns)

    print(odds, '\n')
    driver;close()

Answer 1

如果您使用 python>3.5，ThreadPoolExecutor 會做得很完美。

from selenium import webdriver
import statistics as stat
import numpy as np
from concurrent.futures import ThreadPoolExecutor

driver = webdriver.Firefox()
url = 'https://www.coteur.com/cotes-foot.php'
driver.get(url)

#Store url associated with the soccer games
url_links = []
for i in driver.find_elements_by_xpath('//a[contains(@href, "match/cotes-")]'):
    url_links.append(i.get_attribute('href'))

driver.close()
print(len(url_links), '\n')

def sraper(url_link):
    driver = webdriver.Firefox()
    driver.get(url_link)

    #Store odds into table
    odds = []
    header = []
    for i in driver.find_elements_by_xpath('//button[contains(@class, "btn btn-default btn-xs btncote")]'):
        odds.append(i.text)

    for i in driver.find_elements_by_xpath('//th[contains(@width, "20%")]'):
        header.append(i.text)

    rows = int(len(odds)/3) 
    columns = 3
    odds = [float(i) for i in odds]
    odds = np.array(odds)
    header = np.array(header)
    odds = odds.reshape(rows, columns)

    print(odds, '\n')
    driver;close()
    
with ThreadPoolExecutor(max_workers=8) as executor:
    executor.map(sraper, url_links)