Python爬虫：如何获取城市租房信息？

时间：2023-05-07 22:00

思路：先单线程爬虫，测试可以成功爬取之后再优化为多线程，最后存入数据库

以爬取郑州市租房信息为例

注意：本实战项目仅以学习为目的，为避免给网站造成太大压力，请将代码中的num修改成较小的数字，并将线程改小

一、单线程爬虫

# 用session取代requests# 解析库使用bs4# 并发库使用concurrentimport requests# from lxml import etree    # 使用xpath解析from bs4 import BeautifulSoupfrom urllib import parseimport reimport time headers = {    'referer': 'https://zz.zu.fang.com/',    'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36',    'cookie': 'global_cookie=ffzvt3kztwck05jm6twso2wjw18kl67hqft; city=zz; integratecover=1; __utma=147393320.427795962.1613371106.1613371106.1613371106.1; __utmc=147393320; __utmz=147393320.1613371106.1.1.utmcsr=zz.fang.com|utmccn=(referral)|utmcmd=referral|utmcct=/; __utmt_t0=1; __utmt_t1=1; __utmt_t2=1; ASP.NET_SessionId=aamzdnhzct4i5mx3ak4cyoyp; Rent_StatLog=23d82b94-13d6-4601-9019-ce0225c092f6; Captcha=61584F355169576F3355317957376E4F6F7552365351342B7574693561766E63785A70522F56557370586E3376585853346651565256574F37694B7074576B2B34536C5747715856516A4D3D; g_sourcepage=zf_fy%5Elb_pc; unique_cookie=U_ffzvt3kztwck05jm6twso2wjw18kl67hqft*6; __utmb=147393320.12.10.1613371106'}data={    'agentbid':''} session = requests.session()session.headers = headers # 获取页面def getHtml(url):    try:        re = session.get(url)        re.encoding = re.apparent_encoding        return re.text    except:        print(re.status_code) # 获取页面总数量def getNum(text):    soup = BeautifulSoup(text, 'lxml')    txt = soup.select('.fanye .txt')[0].text    # 取出“共**页”中间的数字    num = re.search(r'd+', txt).group(0)    return num # 获取详细链接def getLink(tex):    soup=BeautifulSoup(text,'lxml')    links=soup.select('.title a')    for link in links:        href=parse.urljoin('https://zz.zu.fang.com/',link['href'])        hrefs.append(href) # 解析页面def parsePage(url):    res=session.get(url)    if res.status_code==200:        res.encoding=res.apparent_encoding        soup=BeautifulSoup(res.text,'lxml')        try:            title=soup.select('div .title')[0].text.strip().replace(' ','')            price=soup.select('div .trl-item')[0].text.strip()            block=soup.select('.rcont #agantzfxq_C02_08')[0].text.strip()            building=soup.select('.rcont #agantzfxq_C02_07')[0].text.strip()            try:                address=soup.select('.trl-item2 .rcont')[2].text.strip()            except:                address=soup.select('.trl-item2 .rcont')[1].text.strip()            detail1=soup.select('.clearfix')[4].text.strip().replace('


',',').replace('
','')            detail2=soup.select('.clearfix')[5].text.strip().replace('


',',').replace('
','')            detail=detail1+detail2            name=soup.select('.zf_jjname')[0].text.strip()            buserid=re.search('buserid: '(d+)'',res.text).group(1)            phone=getPhone(buserid)            print(title,price,block,building,address,detail,name,phone)            house = (title, price, block, building, address, detail, name, phone)            info.append(house)        except:            pass    else:        print(re.status_code,re.text) # 获取代理人号码def getPhone(buserid):    url='https://zz.zu.fang.com/RentDetails/Ajax/GetAgentVirtualMobile.aspx'    data['agentbid']=buserid    res=session.post(url,data=data)    if res.status_code==200:        return res.text    else:        print(res.status_code)        return if __name__ == '__main__':    start_time=time.time()    hrefs=[]    info=[]    init_url = 'https://zz.zu.fang.com/house/'    num=getNum(getHtml(init_url))    for i in range(0,num):        url = f'https://zz.zu.fang.com/house/i3{i+1}/'        text=getHtml(url)        getLink(text)    print(hrefs)    for href in hrefs:        parsePage(href)     print("共获取%d条数据"%len(info))    print("共耗时{}".format(time.time()-start_time))    session.close()

二、优化为多线程爬虫

# 用session取代requests# 解析库使用bs4# 并发库使用concurrentimport requests# from lxml import etree    # 使用xpath解析from bs4 import BeautifulSoupfrom concurrent.futures import ThreadPoolExecutorfrom urllib import parseimport reimport time headers = {    'referer': 'https://zz.zu.fang.com/',    'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36',    'cookie': 'global_cookie=ffzvt3kztwck05jm6twso2wjw18kl67hqft; integratecover=1; city=zz; keyWord_recenthousezz=%5b%7b%22name%22%3a%22%e6%96%b0%e5%af%86%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a014868%2f%22%2c%22sort%22%3a1%7d%2c%7b%22name%22%3a%22%e4%ba%8c%e4%b8%83%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a014864%2f%22%2c%22sort%22%3a1%7d%2c%7b%22name%22%3a%22%e9%83%91%e4%b8%9c%e6%96%b0%e5%8c%ba%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a0842%2f%22%2c%22sort%22%3a1%7d%5d; __utma=147393320.427795962.1613371106.1613558547.1613575774.5; __utmc=147393320; __utmz=147393320.1613575774.5.4.utmcsr=zz.fang.com|utmccn=(referral)|utmcmd=referral|utmcct=/; ASP.NET_SessionId=vhrhxr1tdatcc1xyoxwybuwv; g_sourcepage=zf_fy%5Elb_pc; Captcha=4937566532507336644D6557347143746B5A6A6B4A7A48445A422F2F6A51746C67516F31357446573052634562725162316152533247514250736F72775566574A2B33514357304B6976343D; __utmt_t0=1; __utmt_t1=1; __utmt_t2=1; __utmb=147393320.9.10.1613575774; unique_cookie=U_0l0d1ilf1t0ci2rozai9qi24k1pkl9lcmrs*4'}data={    'agentbid':''} session = requests.session()session.headers = headers # 获取页面def getHtml(url):    res = session.get(url)    if res.status_code==200:        res.encoding = res.apparent_encoding        return res.text    else:        print(res.status_code) # 获取页面总数量def getNum(text):    soup = BeautifulSoup(text, 'lxml')    txt = soup.select('.fanye .txt')[0].text    # 取出“共**页”中间的数字    num = re.search(r'd+', txt).group(0)    return num # 获取详细链接def getLink(url):    text=getHtml(url)    soup=BeautifulSoup(text,'lxml')    links=soup.select('.title a')    for link in links:        href=parse.urljoin('https://zz.zu.fang.com/',link['href'])        hrefs.append(href) # 解析页面def parsePage(url):    res=session.get(url)    if res.status_code==200:        res.encoding=res.apparent_encoding        soup=BeautifulSoup(res.text,'lxml')        try:            title=soup.select('div .title')[0].text.strip().replace(' ','')            price=soup.select('div .trl-item')[0].text.strip()            block=soup.select('.rcont #agantzfxq_C02_08')[0].text.strip()            building=soup.select('.rcont #agantzfxq_C02_07')[0].text.strip()            try:                address=soup.select('.trl-item2 .rcont')[2].text.strip()            except:                address=soup.select('.trl-item2 .rcont')[1].text.strip()            detail1=soup.select('.clearfix')[4].text.strip().replace('


',',').replace('
','')            detail2=soup.select('.clearfix')[5].text.strip().replace('


',',').replace('
','')            detail=detail1+detail2            name=soup.select('.zf_jjname')[0].text.strip()            buserid=re.search('buserid: '(d+)'',res.text).group(1)            phone=getPhone(buserid)            print(title,price,block,building,address,detail,name,phone)            house = (title, price, block, building, address, detail, name, phone)            info.append(house)        except:            pass    else:        print(re.status_code,re.text) # 获取代理人号码def getPhone(buserid):    url='https://zz.zu.fang.com/RentDetails/Ajax/GetAgentVirtualMobile.aspx'    data['agentbid']=buserid    res=session.post(url,data=data)    if res.status_code==200:        return res.text    else:        print(res.status_code)        return if __name__ == '__main__':    start_time=time.time()    hrefs=[]    info=[]    init_url = 'https://zz.zu.fang.com/house/'    num=getNum(getHtml(init_url))    with ThreadPoolExecutor(max_workers=5) as t:        for i in range(0,num):            url = f'https://zz.zu.fang.com/house/i3{i+1}/'            t.submit(getLink,url)    print("共获取%d个链接"%len(hrefs))    print(hrefs)    with ThreadPoolExecutor(max_workers=30) as t:        for href in hrefs:            t.submit(parsePage,href)    print("共获取%d条数据"%len(info))    print("耗时{}".format(time.time()-start_time))    session.close()

三、使用asyncio进一步优化

# 用session取代requests# 解析库使用bs4# 并发库使用concurrentimport requests# from lxml import etree    # 使用xpath解析from bs4 import BeautifulSoupfrom concurrent.futures import ThreadPoolExecutorfrom urllib import parseimport reimport timeimport asyncio headers = {    'referer': 'https://zz.zu.fang.com/',    'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36',    'cookie': 'global_cookie=ffzvt3kztwck05jm6twso2wjw18kl67hqft; integratecover=1; city=zz; keyWord_recenthousezz=%5b%7b%22name%22%3a%22%e6%96%b0%e5%af%86%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a014868%2f%22%2c%22sort%22%3a1%7d%2c%7b%22name%22%3a%22%e4%ba%8c%e4%b8%83%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a014864%2f%22%2c%22sort%22%3a1%7d%2c%7b%22name%22%3a%22%e9%83%91%e4%b8%9c%e6%96%b0%e5%8c%ba%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a0842%2f%22%2c%22sort%22%3a1%7d%5d; __utma=147393320.427795962.1613371106.1613558547.1613575774.5; __utmc=147393320; __utmz=147393320.1613575774.5.4.utmcsr=zz.fang.com|utmccn=(referral)|utmcmd=referral|utmcct=/; ASP.NET_SessionId=vhrhxr1tdatcc1xyoxwybuwv; g_sourcepage=zf_fy%5Elb_pc; Captcha=4937566532507336644D6557347143746B5A6A6B4A7A48445A422F2F6A51746C67516F31357446573052634562725162316152533247514250736F72775566574A2B33514357304B6976343D; __utmt_t0=1; __utmt_t1=1; __utmt_t2=1; __utmb=147393320.9.10.1613575774; unique_cookie=U_0l0d1ilf1t0ci2rozai9qi24k1pkl9lcmrs*4'}data={    'agentbid':''} session = requests.session()session.headers = headers # 获取页面def getHtml(url):    res = session.get(url)    if res.status_code==200:        res.encoding = res.apparent_encoding        return res.text    else:        print(res.status_code) # 获取页面总数量def getNum(text):    soup = BeautifulSoup(text, 'lxml')    txt = soup.select('.fanye .txt')[0].text    # 取出“共**页”中间的数字    num = re.search(r'd+', txt).group(0)    return num # 获取详细链接def getLink(url):    text=getHtml(url)    soup=BeautifulSoup(text,'lxml')    links=soup.select('.title a')    for link in links:        href=parse.urljoin('https://zz.zu.fang.com/',link['href'])        hrefs.append(href) # 解析页面def parsePage(url):    res=session.get(url)    if res.status_code==200:        res.encoding=res.apparent_encoding        soup=BeautifulSoup(res.text,'lxml')        try:            title=soup.select('div .title')[0].text.strip().replace(' ','')            price=soup.select('div .trl-item')[0].text.strip()            block=soup.select('.rcont #agantzfxq_C02_08')[0].text.strip()            building=soup.select('.rcont #agantzfxq_C02_07')[0].text.strip()            try:                address=soup.select('.trl-item2 .rcont')[2].text.strip()            except:                address=soup.select('.trl-item2 .rcont')[1].text.strip()            detail1=soup.select('.clearfix')[4].text.strip().replace('


',',').replace('
','')            detail2=soup.select('.clearfix')[5].text.strip().replace('


',',').replace('
','')            detail=detail1+detail2            name=soup.select('.zf_jjname')[0].text.strip()            buserid=re.search('buserid: '(d+)'',res.text).group(1)            phone=getPhone(buserid)            print(title,price,block,building,address,detail,name,phone)            house = (title, price, block, building, address, detail, name, phone)            info.append(house)        except:            pass    else:        print(re.status_code,re.text) # 获取代理人号码def getPhone(buserid):    url='https://zz.zu.fang.com/RentDetails/Ajax/GetAgentVirtualMobile.aspx'    data['agentbid']=buserid    res=session.post(url,data=data)    if res.status_code==200:        return res.text    else:        print(res.status_code)        return # 获取详细链接的线程池async def Pool1(num):    loop=asyncio.get_event_loop()    task=[]    with ThreadPoolExecutor(max_workers=5) as t:        for i in range(0,num):            url = f'https://zz.zu.fang.com/house/i3{i+1}/'            task.append(loop.run_in_executor(t,getLink,url)) # 解析页面的线程池async def Pool2(hrefs):    loop=asyncio.get_event_loop()    task=[]    with ThreadPoolExecutor(max_workers=30) as t:        for href in hrefs:            task.append(loop.run_in_executor(t,parsePage,href)) if __name__ == '__main__':    start_time=time.time()    hrefs=[]    info=[]    task=[]    init_url = 'https://zz.zu.fang.com/house/'    num=getNum(getHtml(init_url))    loop = asyncio.get_event_loop()    loop.run_until_complete(Pool1(num))    print("共获取%d个链接"%len(hrefs))    print(hrefs)    loop.run_until_complete(Pool2(hrefs))    loop.close()    print("共获取%d条数据"%len(info))    print("耗时{}".format(time.time()-start_time))    session.close()

四、存入Mysql数据库

（一）建表

from sqlalchemy import create_enginefrom sqlalchemy import String, Integer, Column, Textfrom sqlalchemy.orm import sessionmakerfrom sqlalchemy.orm import scoped_session  # 多线程爬虫时避免出现线程安全问题from sqlalchemy.ext.declarative import declarative_base BASE = declarative_base()  # 实例化engine = create_engine(    "mysql+pymysql://root:root@127.0.0.1:3306/pytest?charset=utf8",    max_overflow=300,  # 超出连接池大小最多可以创建的连接    pool_size=100,  # 连接池大小    echo=False,  # 不显示调试信息)  class House(BASE):    __tablename__ = 'house'    id = Column(Integer, primary_key=True, autoincrement=True)    title=Column(String(200))    price=Column(String(200))    block=Column(String(200))    building=Column(String(200))    address=Column(String(200))    detail=Column(Text())    name=Column(String(20))    phone=Column(String(20))  BASE.metadata.create_all(engine)Session = sessionmaker(engine)sess = scoped_session(Session)

（二）将数据存入数据库中

# 用session取代requests# 解析库使用bs4# 并发库使用concurrentimport requestsfrom bs4 import BeautifulSoupfrom concurrent.futures import ThreadPoolExecutorfrom urllib import parsefrom mysqldb import sess, Houseimport reimport timeimport asyncio headers = {    'referer': 'https://zz.zu.fang.com/',    'user-agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/86.0.4240.198 Safari/537.36',    'cookie': 'global_cookie=ffzvt3kztwck05jm6twso2wjw18kl67hqft; integratecover=1; city=zz; __utmc=147393320; ASP.NET_SessionId=vhrhxr1tdatcc1xyoxwybuwv; __utma=147393320.427795962.1613371106.1613575774.1613580597.6; __utmz=147393320.1613580597.6.5.utmcsr=zz.fang.com|utmccn=(referral)|utmcmd=referral|utmcct=/; __utmt_t0=1; __utmt_t1=1; __utmt_t2=1; Rent_StatLog=c158b2a7-4622-45a9-9e69-dcf6f42cf577; keyWord_recenthousezz=%5b%7b%22name%22%3a%22%e4%ba%8c%e4%b8%83%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a014864%2f%22%2c%22sort%22%3a1%7d%2c%7b%22name%22%3a%22%e9%83%91%e4%b8%9c%e6%96%b0%e5%8c%ba%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a0842%2f%22%2c%22sort%22%3a1%7d%2c%7b%22name%22%3a%22%e7%bb%8f%e5%bc%80%22%2c%22detailName%22%3a%22%22%2c%22url%22%3a%22%2fhouse-a014871%2f%22%2c%22sort%22%3a1%7d%5d; g_sourcepage=zf_fy%5Elb_pc; Captcha=6B65716A41454739794D666864397178613772676C75447A4E746C657144775A347A6D42554F446532357649643062344F6976756E563450554E59594B7833712B413579506C4B684958343D; unique_cookie=U_0l0d1ilf1t0ci2rozai9qi24k1pkl9lcmrs*14; __utmb=147393320.21.10.1613580597'}data={    'agentbid':''} session = requests.session()session.headers = headers # 获取页面def getHtml(url):    res = session.get(url)    if res.status_code==200:        res.encoding = res.apparent_encoding        return res.text    else:        print(res.status_code) # 获取页面总数量def getNum(text):    soup = BeautifulSoup(text, 'lxml')    txt = soup.select('.fanye .txt')[0].text    # 取出“共**页”中间的数字    num = re.search(r'd+', txt).group(0)    return num # 获取详细链接def getLink(url):    text=getHtml(url)    soup=BeautifulSoup(text,'lxml')    links=soup.select('.title a')    for link in links:        href=parse.urljoin('https://zz.zu.fang.com/',link['href'])        hrefs.append(href) # 解析页面def parsePage(url):    res=session.get(url)    if res.status_code==200:        res.encoding=res.apparent_encoding        soup=BeautifulSoup(res.text,'lxml')        try:            title=soup.select('div .title')[0].text.strip().replace(' ','')            price=soup.select('div .trl-item')[0].text.strip()            block=soup.select('.rcont #agantzfxq_C02_08')[0].text.strip()            building=soup.select('.rcont #agantzfxq_C02_07')[0].text.strip()            try:                address=soup.select('.trl-item2 .rcont')[2].text.strip()            except:                address=soup.select('.trl-item2 .rcont')[1].text.strip()            detail1=soup.select('.clearfix')[4].text.strip().replace('


',',').replace('
','')            detail2=soup.select('.clearfix')[5].text.strip().replace('


',',').replace('
','')            detail=detail1+detail2            name=soup.select('.zf_jjname')[0].text.strip()            buserid=re.search('buserid: '(d+)'',res.text).group(1)            phone=getPhone(buserid)            print(title,price,block,building,address,detail,name,phone)            house = (title, price, block, building, address, detail, name, phone)            info.append(house)            try:                house_data=House(                    title=title,                    price=price,                    block=block,                    building=building,                    address=address,                    detail=detail,                    name=name,                    phone=phone                )                sess.add(house_data)                sess.commit()            except Exception as e:                print(e)    # 打印错误信息                sess.rollback()  # 回滚        except:            pass    else:        print(re.status_code,re.text) # 获取代理人号码def getPhone(buserid):    url='https://zz.zu.fang.com/RentDetails/Ajax/GetAgentVirtualMobile.aspx'    data['agentbid']=buserid    res=session.post(url,data=data)    if res.status_code==200:        return res.text    else:        print(res.status_code)        return # 获取详细链接的线程池async def Pool1(num):    loop=asyncio.get_event_loop()    task=[]    with ThreadPoolExecutor(max_workers=5) as t:        for i in range(0,num):            url = f'https://zz.zu.fang.com/house/i3{i+1}/'            task.append(loop.run_in_executor(t,getLink,url)) # 解析页面的线程池async def Pool2(hrefs):    loop=asyncio.get_event_loop()    task=[]    with ThreadPoolExecutor(max_workers=30) as t:        for href in hrefs:            task.append(loop.run_in_executor(t,parsePage,href)) if __name__ == '__main__':    start_time=time.time()    hrefs=[]    info=[]    task=[]    init_url = 'https://zz.zu.fang.com/house/'    num=getNum(getHtml(init_url))    loop = asyncio.get_event_loop()    loop.run_until_complete(Pool1(num))    print("共获取%d个链接"%len(hrefs))    print(hrefs)    loop.run_until_complete(Pool2(hrefs))    loop.close()    print("共获取%d条数据"%len(info))    print("耗时{}".format(time.time()-start_time))    session.close()