Python爬虫练习（三）多线程爬取NCBI genome搜索结

由于ncbi的genome搜索结果的页面对应的翻页操作发送的请求是个post请求，并且其参数众多不太好用requests模块直接请求到页面，因此直接用selenium模拟翻页操作获得每页源码数据即可，由于结果比较多，因此考虑使用多线程/异步/多进程等等，这里用的是线程池的操作，我在这里调用8个线程，针对是哺乳类基因组页面的搜索结果，还是比较快的，一共22页，每个步骤/参数/函数的作用都标注出来了。

from lxml import etree
from selenium import webdriver
from multiprocessing.dummy import Pool
from functools import partial 
import os
import requests
# 实现无可视化界面
from selenium.webdriver.chrome.options import Options
# 实现规避检测
from selenium.webdriver import ChromeOptions


def setoption():
    """
    谷歌浏览器常规反反爬的参数设置
    """
    # 实现无可视化界面的操作
    chrome_options = Options()
    chrome_options.add_experimental_option(
        'excludeSwitches', ['enable-logging'])
    chrome_options.add_argument("--headless")
    chrome_options.add_argument("--disable-gpu")
    # 实现规避检测
    option = ChromeOptions()
    option.add_experimental_option("excludeSwitches",
                                   ["enable-automation"])
    return chrome_options, option


def getonepage(pagetext, filepath):
    """
    得到一页所有物种对应详情页的物种名+基因组信息统计，
    并且写入到文件当中。
    """
    tree = etree.HTML(pagetext)
    initurl = "https://www.ncbi.nlm.nih.gov"
    div_list = tree.xpath(
        '//*[@id="maincontent"]/div/div[5]/div[@class="rprt"]')
    headers = {
        "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 \
            (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36"
    }
    for div in div_list:
        detail_url = initurl + div.xpath('.//p/a/@href')[0]
        response = requests.get(detail_url, headers=headers)
        response = response.content.decode()
        detail_html = bytes(bytearray(response, encoding='utf-8'))
        detail_tree = etree.HTML(detail_html)
        name = detail_tree.xpath(
            '//*[@id="maincontent"]/div/div[5]/div/div[2]/table[1]//tr//span[1]/text()')[0]
        summary = "".join(detail_tree.xpath(
            '//*[@id="mmtest1"]/div/div/div/table//tr//text()'))
        print(name, summary, sep="\n")
        with open(filepath, "a", encoding="utf-8") as fp:
            fp.write(name+"\n"+summary+"\n")


def mainprocess(chrome_options, option, executable_path, filepath, thread=4):
    """
    开启selenium无头浏览器，先得到每一页的源码数据存储，
    然后用每页源码数据作为参数，进行多线程搭建。
    """
    # 让selenium规避被检测到的风险
    bro = webdriver.Chrome(executable_path=executable_path,
                           chrome_options=chrome_options,
                           options=option)
    bro.get("https://www.ncbi.nlm.nih.gov/genome/?term=txid40674%5BOrganism%3Aexp%5D")
    # 生成用于多线程使用的参数列表
    pagetext_list = []
    # 获取当前页源码数据
    pagetext = bro.page_source
    print("Append page1 to the queue.")
    pagetext_list.append(pagetext)
    # 获取全部页码总数
    allpagetree = etree.HTML(pagetext)
    allpage = int(allpagetree.xpath('//*[@id="pageno2"]/@last')[0])
    # 将每页的源码数据加入多线程参数列表
    for pagenum in range(2, allpage+1):
        next_btn = bro.find_element_by_xpath(
            "/html/body/div[1]/div[1]/form/div[1]/div[4]/div/div[7]/div/a[1]")
        next_btn.click()
        pagetext = bro.page_source
        print(f"Append page{pagenum} to the queue.")
        pagetext_list.append(pagetext)
    # 检测是否存在之前的文件
    if os.path.isfile(filepath):
        os.remove(filepath)
    # 多线程使用
    pool = Pool(thread)
    # param = {pagetext: pagetext_list, filepath: filepath}
    pool.map(partial(getonepage, filepath=filepath), pagetext_list)
    pool.close()
    pool.join()
    bro.quit()


if __name__ == "__main__":
    chrome_options, option = setoption()
    mainprocess(chrome_options, option,
                r"chromedriver.exe", "genomeinfo.txt", 8)