利用python爬取电影天堂搜索部分，并利用pyqt可视化

爬虫部分

def crawls_home(film_name):
    name = film_name.encode('gb2312')
    name = str(name)[1:]
    name = re.sub(r'\\x','%',name)
    name = eval(name).upper() #关键字处理
    url = "http://s.ygdy8.com/plus/so.php?typeid=1&keyword={}".format(name)
    html = request_url(url)
    soup = BeautifulSoup(html,'lxml')
    div = soup.find('div',class_="co_content8")
    if div is None:
        return ['没有搜到相应电影']
    film_list = div.find_all('table',border="0",width="100%")
    return film_list

像电影天堂这种的爬虫应该算是很简单的了，这里需要注意的是电影天堂对输入的关键字采用的是gb2312编码，所以需要要对关键字做一些处理

 

可视化部分

主要是利用pyqt5进行简单粗糙的可视化

通过捕捉回车键或者点击确认弹出搜索电影的详细列表

可以点击超链接弹出详细介绍

通过点击超链接调用迅雷实现下载

完整代码如下：

crawlsdytt.py:

import requests
import re
from bs4 import BeautifulSoup

header = {"User-Agent": "Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) "
                            "Chrome/57.0.2987.110 Safari/537.36"}

#film_name = '猫和老鼠'

def request_url(url):  
    while True:
        try:
            r = requests.get(url,headers=header)
            if r.status_code == 200:
                r.encoding = 'gbk'
                return r.text
        except:
            pass

def crawls_home(film_name):
    name = film_name.encode('gb2312')
    name = str(name)[1:]
    name = re.sub(r'\\x','%',name)
    name = eval(name).upper()
    url = "http://s.ygdy8.com/plus/so.php?typeid=1&keyword={}".format(name)
    html = request_url(url)
    soup = BeautifulSoup(html,'lxml')
    div = soup.find('div',class_="co_content8")
    if div is None:
        return ['没有搜到相应电影']
    film_list = div.find_all('table',border="0",width="100%")
    return film_list

def crawls_single(url):
    html = request_url(url)
    soup = BeautifulSoup(html,'lxml')
    div = soup.find('div',id="Zoom")
    node = div.p
    href = re.search(r'href=\"(.*?)\"',str(div)).group(1)
    n_node = '\n\n<a href="{0}">{0}</a>'.format(href)
    node = str(node) + n_node
    return node

qqtt.py:

from PyQt5.QtWidgets import QApplication,QLabel,QWidget,QVBoxLayout,QGridLayout,QPushButton
from PyQt5.QtWidgets import QMainWindow,QScrollArea
from PyQt5.QtCore import *
from PyQt5.Qt import QLineEdit
from crawlsdytt import *
import sys
import re
import os

class WindowDemo(QWidget):
    close_signal = pyqtSignal(object)
    def __init__(self):
        super(WindowDemo,self).__init__()
        self.m = MainWindow()
        self.title = QLabel('电影名称:')
        self.textbox = QLineEdit()
        self.textboxValue = ''
        self.button = QPushButton('确认', self)
        self.button.clicked.connect(self.on_click)

        self.grid = QGridLayout()
        self.grid.setSpacing(10)
        self.grid.addWidget(self.title, 1, 0)
        self.grid.addWidget(self.textbox, 1, 1)
        self.grid.addWidget(self.button,1,2)
        self.setLayout(self.grid)
        self.setWindowTitle("电影天堂")
        
    @pyqtSlot()
    def on_click(self):
        film_name = self.textbox.text()        
        self.m.ui(film_name)

    def keyPressEvent(self, event):
        if(str(event.key()) == '16777220'):
            self.on_click()

        
class Content(QWidget):
    
    def __init__(self):
        super(Content,self).__init__()

    def show_label(self,url):
        print(url)
        textboxValue = crawls_single(url)
        print(str(textboxValue))
        self.title = QLabel(self)
        self.title.setText(str(textboxValue))
        self.title.setWordWrap(True)
        self.title.linkActivated.connect(self.link_clicked)
        self.setWindowTitle("详细信息")
        vbox = QVBoxLayout()
        vbox.addWidget(self.title)
        self.setLayout(vbox)
        self.show()

    def link_clicked(self):
        ss = self.title.text()
        print(ss)
        link = re.search(r'href=\"(.*?)\"',ss).group(1)
        # 启动迅雷
        os.chdir(r'D:\Program Files (x86)\Thunder Network\Thunder\Program')
        os.system("Thunder.exe -StartType:DesktopIcon \"%s\"" % link)

class MainWindow(QMainWindow):
    def __init__(self):
        super(QMainWindow,self).__init__()

    def ui(self,film_name):
        if film_name is None or film_name.strip() == '':
            return
        else:
            film_list = crawls_home(film_name)
        w = QWidget()
        self.setCentralWidget(w)
        self.topFiller = QWidget()
        self.topFiller.setMinimumSize(800, 2000)
        self.qlabel_dict = {}
        self.c = {}
        for i in range(len(film_list)):
            self.c[i] = Content()
            self.qlabel_dict[i] = QLabel(self.topFiller)
            self.qlabel_dict[i].setText(str(film_list[i]))
            self.qlabel_dict[i].setWordWrap(True)
            self.qlabel_dict[i].linkActivated.connect(self.link_clicked(i))
            self.qlabel_dict[i].move(10,i*300)
        self.scroll = QScrollArea()
        self.scroll.setWidget(self.topFiller)
 
        self.vbox = QVBoxLayout()
        self.vbox.addWidget(self.scroll)
        
        w.setLayout(self.vbox)
 
        self.statusBar().showMessage("电影列表")
        self.setWindowTitle("详细列表")
        self.show()

    def link_clicked(self,i):
        def link():
            ss = self.qlabel_dict[i].text()
            #print(ss)
            url ="http://www.ygdy8.com" + re.search(r'href=\"(.*?)\"',str(ss)).group(1)
            self.c[i].show_label(url)
            return url
        return link

if __name__ == '__main__':
    app = QApplication(sys.argv)
    win = WindowDemo()
    win.show()
    sys.exit(app.exec_())