《进击的虫师》百度贴吧通用抓图脚本

多进程优势:单个进程的崩溃,不会影响其它进程 随之而来的问题是,进程之间,资源不共享,信息不共享,所以进程通讯的问题,是实现多进程协作,必须解决的问题

为解决进程间的通讯,人们常用的方法是 --> 创建一个中间人(队列),作为他们交流的中介...

###以爬取某贴吧帖子中的所有图片为例:

大致需要四步

一. 获取所有的 帖子的目录

二. 根据帖子的目录,获取 帖子详情页

三. 根据 帖子详情页 的源码,获取内嵌的 图片url信息

四. 根据 图片url信息,下载图片到本地

from multiprocessing import Process, Queue
import time
from time import sleep
import requests
from lxml import etree
import os, sys
import re

class BaiduTb(object):
	def __init__(self, tb_name):
		self.start_url = "https://tieba.baidu.com/f?ie=utf-8&kw=" + tb_name
		self.q_list = Queue()
		self.q_detail = Queue()
		self.q_image = Queue()
		self.headers = {
			"User-Agent": "Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0) "
		}
		self.p1 = None
		self.p2 = None
		self.p3 = None
		self.p4 = None
	# 获取响应内容
	def get_response_content(self, url):
		response = requests.get(url, headers = self.headers)
		print("-->",response)
		return response.content
	# 获取下一个网页url
	def get_next_url(self, now_url):
		content = self.get_response_content(now_url)
		# 将响应信息转换为element对象
		html = etree.HTML(content)
		# 通过xpath获取下一页的链接
		next_url = "http:" + html.xpath('//a[@class="next pagination-item "]/@href')[0]
		return next_url

	def get_list_page_urls(self, q_list):
		while True:
			try:
				sleep(1)
				q_list.put(self.start_url)
				self.start_url = self.get_next_url(self.start_url)
			except:
				print("主程序..")

	def get_detail_page_urls(self, q_list, q_detail):
		while True:
			sleep(1)
			try:
				if not q_list.empty():
					list_url = q_list.get(True)
				else:
					continue

			except:
				print("第二进程已完成..")

			content = self.get_response_content(list_url)
			# 将响应信息转换为element对象
			html = etree.HTML(content)

			# 获取 标题 和url 列表
			detail_titles_list = html.xpath('//*[@id="thread_list"]//div/div[2]/div[1]/div[1]/a/text()')
			detail_urls_list = html.xpath('//*[@id="thread_list"]//div/div[2]/div[1]/div[1]/a/@href')

			# 将"标题"和 url 增加到 详情页队列中
			for i in range(len(detail_titles_list)):
				detail_title = detail_titles_list[i]
				detail_url = "https://tieba.baidu.com" + detail_urls_list[i]
				temp_tuple = (detail_title, detail_url)
				q_detail.put(temp_tuple)

	def get_image_urls (self, q_detail, q_image):
		while True:
			sleep(3)
			try:
				detail = q_detail.get()
				detail_title = detail[0]
				detail_url = detail[1]
			except Exception as e:
				print("第三进程已完成")
			# 获取详情页所有图片的url
			content = self.get_response_content(detail_url)
			# 将响应信息转换为element对象
			html = etree.HTML(content)
			# 通过xpath获取下一页的链接
			image_urls = html.xpath('//cc//img/@src')
			for i in range(len(image_urls)):
				image_url = image_urls[i]
				temp_image_info = (detail_title, image_url)
				q_image.put(temp_image_info)

	def save_image(self, q_image):

		while True:
			sleep(3)
			try:
				if not q_image.empty():
					image_info = q_image.get()
					image_name = re.match(r".*(.{10})", image_info[1]).group(1)
					print("图片名称为:", image_name, "图片地址为:", image_info[1], "帖子标题为:", image_info[0])
					# 尝试根据帖子名称,创建新文件夹
					try:
						os.mkdir("./%s"%(image_info[0]))
					except:
						pass
					new_file_path = "./%s/%s"%(image_info[0],image_name)
					with open(new_file_path, "wb") as f:
						data = self.get_response_content(image_info[1])
						f.write(data)
					if (self.q_list.empty()) and (self.q_detail.empty()) and (self.q_image.empty()):
						exit()
			except:
				print("第四进程已完成")

	def run(self):
		print("开始执行")

		self.p1 = Process(target = self.get_list_page_urls, args=(self.q_list,))
		self.p2 = Process(target = self.get_detail_page_urls, args=(self.q_list, self.q_detail,))	
		self.p3 = Process(target = self.get_image_urls, args=(self.q_detail, self.q_image,))
		self.p4 = Process(target = self.save_image, args=(self.q_image,))

		self.p1.start()
		self.p2.start()
		self.p3.start()
		self.p4.start()

		self.p1.join()
		self.p2.join()
		self.p3.join()
		self.p4.join()


def main():
	name = input("请输入贴吧名称:")
	beautiful_girl = BaiduTb(name)
	beautiful_girl.run()


if __name__ == '__main__':
	main()

复制代码

---

多进程和和多线程哪个更好用? 追求资源利用率,考虑多线程 追求程序稳定性,考虑多进程

搞专业爬虫的话,先保证网速够好，再考虑多进程还是多线程~