python 爬取杭州小区挂牌均价

下载chrome驱动

通过chrome浏览器的 设置-帮助-关于Google Chrome 查看你所使用的Chrome版本

驱动可以从这两个地方找:

• 【推荐】https://storage.googleapis.com/chrome-for-testing-public
• http://npm.taobao.org/mirrors/chromedriver

import zipfile
import os
import requests

def un_zip(file_name, to_dir='./'):
    """unzip zip file"""
    zip_file = zipfile.ZipFile(file_name)
    if os.path.isdir(to_dir):
        pass
    else:
        os.mkdir(to_dir)
    for names in zip_file.namelist():
        zip_file.extract(names, to_dir)
    zip_file.close()

def download_driver(url=None, to_dir='./', version=''):
    print('install chrome-driver first')
    if not url:
        url = 'http://npm.taobao.org/mirrors/chromedriver/LATEST_RELEASE'
        if len(version)>0:
            url = 'http://npm.taobao.org/mirrors/chromedriver/LATEST_RELEASE_'+version
    
        version = requests.get(url).content.decode('utf8')
        driver_file = 'http://npm.taobao.org/mirrors/chromedriver/' + version + '/chromedriver_win32.zip'
    else:
        driver_file = url
    r = requests.get(driver_file)
    download_zip = "chromedriver_win32.zip"
    with open(download_zip, "wb") as code:
        code.write(r.content)
    un_zip(download_zip,  to_dir)
    # os.remove(download_zip)
    print('done')

download_driver(url='https://storage.googleapis.com/chrome-for-testing-public/123.0.6312.58/win64/chromedriver-win64.zip', to_dir='./', version='')

数据库交互

把抓取的数据保存到数据库，sqlite数据库是一个db文件，可以用DBeaver工具打开，很方便

import sqlite3

class DBC:
    def __init__(self, dbname):
        self.db = dbname
        self.conn = None

    def __enter__(self):
        self.conn = sqlite3.connect(self.db)
        return self.conn

    def __exit__(self, exc_type, exc_val, exc_tb):
        if exc_type is not None:
            self.conn.rollback()
            print("rollback")
            print(exc_type, exc_val, exc_tb)
        else:
            self.conn.commit()
        self.conn.close()

def insert2db(data):
    with DBC('anjuke.db') as conn:
        cur = conn.cursor()
        cur.executemany("""
        insert into anjuke(community_name,year,address, tags, price, if_down, percent) values
        (?,?,?,?,?,?,?)
        """, data) 

with DBC('anjuke.db') as conn:
    cur = conn.cursor()
    cur.execute("""
    drop table if exists anjuke
    """)
    cur.execute("""
    create table if not exists anjuke(
        id INTEGER primary key  AUTOINCREMENT not null,
        community_name TEXT,
        year TEXT,
        address TEXT,
        tags TEXT,
        price TEXT,
        if_down BOOL,
        percent TEXT
    )
    """)

selenium 爪巴虫

import time
import traceback
from selenium import webdriver
 
browser = webdriver.Chrome()
def process(url):
    browser.get(url)
    html = browser.page_source
    html = BeautifulSoup(html,)
    As = html.find_all("a", {"class": "li-row"})
    if len(As) == 0:
        raise Exception("EMPTY")
    data = []
    for A in As:
        price = A.find("div", {"class": "li-side"})
        price_value = price.find("div", {"class": "community-price"})
        if price_value:
            price_value = price_value.text.strip()
        minus = price.find("span")
        if minus:
            minus = 'propor-green' in minus.attrs['class']
        percent = price.find("span")
        if percent:
            percent = percent.text.strip()
        info = A.find("div", {"class": "li-info"})
        community_name = info.find("div", {"class": "nowrap-min li-community-title"}).text
        year = info.find("span", {"class": "year"})
        if year:
            year = year.text
        advantage =info.find("div", {"class":"prop-tags"})
        if advantage:
            advantage = advantage.text.strip()
        address = info.find("div", {"class": "props nowrap"}).find_all("span")[-1].text
        data.append((community_name,year,address, advantage, price_value, minus, percent))
    insert2db(data)


i = 0
while i < 50:  # 超过50显示重复数据
    i += 1
    url = f'https://hangzhou.anjuke.com/community/o8-p{i}'  # o2,o4,o6,o8不同排序条件
    print(url)
    try:
        process(url)
    except Exception as e:
        traceback.print_exc()
        s = input("check:")  # 抓取过程中，可能需要输入验证码
        i = i-1
    time.sleep(0.5)