利用爬到的数据,基于Django搭建的一个最新电影信息网站:
今天想利用所学知识来爬取电影天堂所有最新电影信息,用到的模块:
requests:用于获取网页信息
re:获取网页中具体想要的信息
Beautifulsoup:便于标签搜索,获取想要信息
threading:使用多线程大幅度缩短爬取时间
queue:使用线程队列来写入文件(实际上我是把数据全部存入了数据库)
大概差不多就这些模块。
欢迎各位大牛指点。
# Author : 'n1celll' import requests import json import re from bs4 import BeautifulSoup import threading import queue,time header = header = { 'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/73.0.3683.86 Safari/537.36'} url = 'http://www.ygdy8.net/html/gndy/dyzz/index.html' def get_page(url): index = requests.get(url, headers=header) index.enconding = 'GBK'# 将编码转为与HTML一致 t = index.text index_soup = BeautifulSoup(t, 'html.parser')# 将获得的网页信息 转成soup对象 all_pages = index_soup.find('select', attrs={'name': 'sldd'}).find_all('option')[-1] # 获得总页数 page = int(all_pages.string) return page def get_data(page): page_url = 'http://www.ygdy8.net/html/gndy/dyzz/list_23_%s.html' % (page) # 获取每一页数据 print(page) # page_url = 'http://www.ygdy8.net/html/gndy/dyzz/list_23_30.html' res = requests.get(page_url, headers=header) res.encoding = 'GBK' # 'gb2312' a = res.text soup = BeautifulSoup(a, 'html.parser') name = soup.find_all('a', attrs={'class': 'ulink'}) # print(name)测试 for i in name: try: moive_name = re.search('《(.*?)(》|】)', i.string).group() # 有两个坑,这个里面有个电影名字不是用的书名号,还有一个电影有两个a标签 except: continue html = 'http://www.ygdy8.net' + i['href'] da = requests.get(html, headers=header) da.encoding = 'GBK' # da.apparent_encoding db = da.text # f = open('test2.txt','w',encoding='utf8') # f.write(a.text) # f.close() dr = BeautifulSoup(db, 'html.parser') span = dr.find('span', attrs={'style': 'FONT-SIZE: 12px'}) if span: dc = span.text.split() data = '' for i in dc: data += i print(data) msg = {} if data: msg['mname'] = moive_name try: show_t = re.search(r'(?<=(◎年代|◎时间|品年代|年代】|播时间|播】:))(.*?)(?=◎|年|【)', data).group() except: show_t = re.search(r'(?<=日期|份:)(.*?)(?=(-|剧))', data).group() msg['mtime'] = show_t try: country = re.search(r'(?<=(◎国家|◎产地|◎地区|◎国别|国家】))(.*?)(?=◎|【类)', data).group() except: try: country = re.search(r'(?<=地区)(.*?)(?=语言)', data).group() except: country = '未知' msg['mcountry'] = country try: time = re.search(r'(?<=◎片长|长度】)(.*?)(?=◎|【)', data).group() except: time = '未知' msg['mtime'] = time try: mtype = re.search( r'(?<=(◎类别|别类型|影类型|◎类型|集类型|◎分类|类型:|类别】|片类型|型】:))(.*?)(?=(◎|级别|【出品|【主演))', data).group() except: try: mtype = re.search(r'(?<=类型:)(.*?)(?=国)', data).group() except: mtype = re.search(r'动作|爱情|战争', data).group() #以上的正则表达式,感觉用的很笨拙,希望有技术大牛提点建议 # with open('test4.txt','a+',encoding='utf8') as f:测试 # f.write('%s: %s,%s,%s,%s ' % (moive_name, country, mtype, time, show_t))测试 q.put('%s: %s,%s,%s,%s,%s ' % (moive_name, country, mtype, time, show_t,html)) q = queue.Queue(maxsize=10000) t_obj = [] lock = threading.Lock()#加上线程锁 # semaphore = threading.BoundedSemaphore(200) def writing(f): # semaphore.acquire() data = q.get() lock.acquire() f.write(data) lock.release() # semaphore.release() # if not q.get(): # f.close() # break # print('写入完成') all_page = get_page(url) f = open('test4.txt', 'w', encoding='utf8') print(all_page+1) for i in range(1,all_page+1): t = threading.Thread(target=get_data,args=(i,)) t.start() t_obj.append(t) for t in t_obj: t.join()#保证所有线程结束后开始写入 print('%s over'%t) while q.qsize():#判断队列里面是否还有元素 w = threading.Thread(target=writing, args=(f,)) w.start() w.join() else: print('写入完成')