本文介绍了一个基于Python的分布式爬虫系统实现方案，通过多进程和网络通信技术，实现了URL管理、网页下载、数据解析及存储等功能。系统采用Master-Worker架构，能够高效抓取网页并解析数据。
DataOutPut.py
#!/usr/bin/python
# -*- coding: UTF-8 -*-
import codecs
import time
class DataOutPut(object):
def __init__(self):
self.filepath = 'baike_%s.html' % (time.strftime("%Y_%m_%d_%H_%M_%S", time.localtime()))
self.output_head(self.filepath)
self.datas = []
def store_data(self, data):
if data is None:
return
self.datas.append(data)
if len(self.datas) > 10:
self.output_html(self.filepath)
def output_head(self, path):
'''
将HTML头写进去
:return:
'''
fout = codecs.open(path, 'w', encoding='utf-8')
fout.write("<html>")
fout.write(r'''<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />''')
fout.write("<body>")
fout.write("<table>")
fout.close()
def output_html(self, path):
'''
将数据写入HTML文件中
:param path: 文件路径
:return:
'''
fout = codecs.open(path, 'a', encoding='utf-8')
for data in self.datas:
fout.write("<tr>")
fout.write("<td>%s</td>" % data['url'])
fout.write("<td>%s</td>" % data['title'])
fout.write("<td>%s</td>" % data['summary'])
fout.write("</tr>")
self.datas = []
fout.close()
def ouput_end(self, path):
'''
输出HTML结束
:param path: 文件存储路径
:return:
'''
fout = codecs.open(path, 'a', encoding='utf-8')
fout.write("</table>")
fout.write("</body>")
fout.write("</html>")
fout.close()
NodeManager.py
#!/usr/bin/python
# -*- coding: UTF-8 -*-
from multiprocessing.managers import BaseManager
import time
from multiprocessing import Process, Queue
from DistributeSpider.ControNode.DataOutPut import DataOutPut
from DistributeSpider.ControNode.URLManager import URLManager
class NodeManager(object):
def start_Manager(self, url_q, result_q):
'''
创建一个分布式管理器
:param url_q: url队列
:param result_q: 结果队列
:return:
'''
# 把创建的两个队列注册在网络上，利用register方法，callable参数关联了Queue对象，
# 将Queue对象在网络中暴露
BaseManager.register('get_task_queue', callable=lambda: url_q)
BaseManager.register('get_result_queue', callable=lambda: result_q)
# 绑定端口8001 设置验证口令 相当于对象的初始化
manager = BaseManager(address=('', 8001), authkey='ctj'.encode('utf-8'))
# 返回manager对象
return manager
def url_manager_proc(self, url_q, conn_q, root_url):
url_manager = URLManager()
url_manager.add_new_url(root_url)
while True:
while url_manager.has_new_url():
# 从URL管理器中获取新的url
new_url = url_manager.get_new_url()
# 将新的URL发送给工作节点
url_q.put(new_url)
print('old_url=', url_manager.old_url_size())
# 加一个判断条件，当爬去2000个链接后就关闭,并保存进度
if (url_manager.old_url_size() > 2000):
# 通知爬行节点工作结束
url_q.put('end')
print('控制节点发起结束通知!')
# 关闭管理节点，同时存储set状态
url_manager.save_progress('new_urls.txt', url_manager.new_urls)
url_manager.save_progress('old_urls.txt', url_manager.old_urls)
return
# 将从result_solve_proc获取到的urls添加到URL管理器之间
try:
urls = conn_q.get()
url_manager.add_new_urls(urls)
except BaseException as e:
time.sleep(0.1) # 延时休息
def result_solve_proc(self, result_q, conn_q, store_q):
while (True):
try:
if not result_q.empty():
# Queue.get(block=True, timeout=None)
content = result_q.get(True)
if content['new_urls'] == 'end':
# 结果分析进程接受通知然后结束
print('结果分析进程接受通知然后结束!')
store_q.put('end')
return
conn_q.put(content['new_urls']) # url为set类型
store_q.put(content['data']) # 解析出来的数据为dict类型
else:
time.sleep(0.1) # 延时休息
except BaseException as e:
time.sleep(0.1) # 延时休息
def store_proc(self, store_q):
output = DataOutPut()
while True:
if not store_q.empty():
data = store_q.get()
if data == 'end':
print('存储进程接受通知然后结束!')
output.ouput_end(output.filepath)
return
output.store_data(data)
else:
time.sleep(0.1)
pass
if __name__ == '__main__':
# 初始化4个队列
url_q = Queue()
result_q = Queue()
store_q = Queue()
conn_q = Queue()
# 创建分布式管理器
node = NodeManager()
manager = node.start_Manager(url_q, result_q)
# 创建URL管理进程 数据提取进程 数据存储进程
url_manager_proc = Process(target=node.url_manager_proc,
args=(url_q, conn_q, 'http://baike.baidu.com/view/284853.htm',))
result_solve_proc = Process(target=node.result_solve_proc, args=(result_q, conn_q, store_q,))
store_proc = Process(target=node.store_proc, args=(store_q,))
# 启动3个进程和分布式管理器
url_manager_proc.start()
result_solve_proc.start()
store_proc.start()
manager.get_server().serve_forever()
URLManager.py
#!/usr/bin/python
# -*- coding: UTF-8 -*-
import pickle
import hashlib
class URLManager(object):
def __init__(self):
self.new_urls = self.load_progress('new_urls.txt') # 未爬取的URl集合
self.old_urls = self.load_progress('old_urls.txt') # 已爬取的URL集合
def has_new_url(self):
return self.new_url_size() != 0
def get_new_url(self):
new_url = self.new_urls.pop()
m = hashlib.md5()
m.update(new_url.encode('utf-8'))
self.old_urls.add(m.hexdigest()[8:-8])
return new_url
def add_new_url(self, url):
if url is None:
return
m = hashlib.md5()
m.update(url.encode('utf-8'))
url_md5 = m.hexdigest()[8:-8]
if url not in self.new_urls and url_md5 not in self.old_urls:
self.new_urls.add(url)
def add_new_urls(self, urls):
if urls is None or len(urls) == 0:
for url in urls:
self.add_new_url(url)
def new_url_size(self):
return len(self.new_urls)
def old_url_size(self):
return len(self.old_urls)
def save_progress(self, path, data):
with open(path, 'wb') as f:
pickle.dump(data, f)
def load_progress(self, path):
print('[+] 从文件中加载进度：%s' % path)
try:
with open(path, 'rb') as f:
tmp = pickle.load(f)
return tmp
except:
print('[!] 无进度条文件,创建：%s' % path)
return set()
HtmlDownloader.py
#!/usr/bin/python
# -*- coding: UTF-8 -*-
import requests
class HtmlDownloader(object):
def download(self, url):
if url is None:
return None
user_agent = 'Mozilla/4.0 (compatible; MSIE 5.5; Windows NT)'
headers = {'User-Agent': user_agent}
r = requests.get(url, headers=headers)
if r.status_code == 200:
r.encoding = 'utf-8'
return r.text
return None
SpiderWork.py
#!/usr/bin/python
# -*- coding: UTF-8 -*-
from multiprocessing.managers import BaseManager
from DistributeSpider.SpiderWork.HtmlDownloader import HtmlDownloader
from DistributeSpider.SpiderWork.HtmlParser import HtmlParser
class SpiderWork(object):
def __init__(self):
# 初始化分布式进程中的工作节点的连接工作
# 实现第一步：使用BaseManager注册获取Queue的方法名称
BaseManager.register('get_task_queue')
BaseManager.register('get_result_queue')
# 实现第二步：连接到服务器:
server_addr = '127.0.0.1'
print(('Connect to server %s...' % server_addr))
# 端口和验证口令注意保持与服务进程设置的完全一致:
self.m = BaseManager(address=(server_addr, 8001), authkey='ctj'.encode('utf-8'))
# 从网络连接:
self.m.connect()
# 实现第三步：获取Queue的对象:
self.task = self.m.get_task_queue()
self.result = self.m.get_result_queue()
# 初始化网页下载器和解析器
self.downloader = HtmlDownloader()
self.parser = HtmlParser()
print('init finish')
def crawl(self):
while (True):
try:
if not self.task.empty():
url = self.task.get()
if url == 'end':
print('控制节点通知爬虫节点停止工作...')
# 接着通知其它节点停止工作
self.result.put({'new_urls': 'end', 'data': 'end'})
return
print('爬虫节点正在解析:%s' % url.encode('utf-8'))
content = self.downloader.download(url)
new_urls, data = self.parser.parser(url, content)
self.result.put({"new_urls": new_urls, "data": data})
except EOFError as e:
print("连接工作节点失败")
return
except Exception as e:
print(e)
print('Crawl fali ')
if __name__ == "__main__":
spider = SpiderWork()
spider.crawl()
HtmlParser.py
#!/usr/bin/python
# -*- coding: UTF-8 -*-
import re
import urllib.parse
from bs4 import BeautifulSoup
class HtmlParser(object):
def parser(self, page_url, html_cont):
'''
用于解析网页内容抽取URL和数据
:param page_url: 下载页面的URL
:param html_cont: 下载的网页内容
:return:返回URL和数据
'''
if page_url is None or html_cont is None:
return
soup = BeautifulSoup(html_cont, 'lxml')
new_urls = self._get_new_urls(page_url, soup)
new_data = self._get_new_data(page_url, soup)
return new_urls, new_data
def _get_new_urls(self, page_url, soup):
'''
抽取新的URL集合
:param page_url: 下载页面的URL
:param soup:soup
:return: 返回新的URL集合
'''
new_urls = set()
# 抽取符合要求的a标签
links = soup.find_all('a', href=re.compile(r'/item/.*'))
for link in links:
# 提取href属性
new_url = link['href']
# 拼接成完整网址
new_full_url = urllib.parse.urljoin(page_url, new_url)
print(new_full_url)
new_urls.add(new_full_url)
return new_urls
def _get_new_data(self, page_url, soup):
'''
抽取有效数据
:param page_url:下载页面的URL
:param soup:
:return:返回有效数据
'''
data = {}
data['url'] = page_url
title = soup.find('dd', class_='lemmaWgt-lemmaTitle-title').find('h1')
data['title'] = title.get_text()
summary = soup.find('div', class_='lemma-summary')
# 获取到tag中包含的所有文版内容包括子孙tag中的内容,并将结果作为Unicode字符串返回
data['summary'] = summary.get_text()
return data
项目的源代码下载 https://github.com/tanjunchen/SpiderProject/tree/master/HouseWorldNewHouse
扫一扫
4184
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
