05 如何在知乎上自动互粉、商品上架提醒、拉勾网职位提醒

任务

image.png

问题拆解

# Windows版的代码请在这下载：https://video.mugglecode.com/net5.py
# 以下为Mac/Linux/可在麻瓜编程在线运行的代码：

from selenium import webdriver
import time

# 运行前先下载 chrome driver,下载地址是：https://sites.google.com/a/chromium.org/chromedriver/downloads，点击【Latest Release: ChromeDriver x.xx】进入下载

def start_chrome():
    driver = webdriver.Chrome(executable_path='./chromedriver') # Windows 需写成'./chromedriver.exe'
    driver.start_client()
    return driver

def find_strangers():
    # btn
    btn_sel = 'div.ContentItem-extra > button.Button--blue'
    elems = driver.find_elements_by_css_selector(btn_sel)
    return elems

def add_fren():
    pass


while True:
    url = 'https://www.zhihu.com/'
    follower_url = 'https://www.zhihu.com/people/xxx/followers' #需替换成你的知乎url，点击【我的主页】→【关注者】可进入该页面
    driver = start_chrome()
    driver.get(url)
    if not driver.get_cookies():
        push()
    time.sleep(20)
        # wait login

    driver.get(follower_url)
    time.sleep(6) # wait for loading page & users
    strangers = find_strangers()
    for s in strangers:
        s.click()
        time.sleep(3)
    print('Done!')
    time.sleep(3000)
# js_execute('xxx.click()')

商品上架提醒这一部分需要考虑的是样品的标签名称。另外pyautogui库里面的alert方法可以提醒。

from selenium.webdriver import Chrome
import pyautogui
import webbrowser
import time


class PageObserver:

    def __init__(self, url, target_sel):
        self.driver = Chrome(executable_path='./chromedriver')
        self.url = url
        self.target_sel = target_sel
        self.request_time = 20
        #self.driver.find_elements_by_css_selector(self.target_sel)

    def is_changed(self):
        self.driver.get(self.url)
        time.sleep(self.request_time)
        oos_el = self.driver.find_elements_by_css_selector(self.target_sel)
        buy_button = self.driver.find_elements_by_id('update-cart')
        # -> []
        print(oos_el)
        print(buy_button)
        if not oos_el and buy_button:
            return True


def alert():
    pyautogui.alert('The bag is available')



# find changes
url = 'https://www.strathberry.com/products/east-west-mini-tri-colour-navy-ruby-vanilla'
target_sel = 'div.oos.swatch-container.swatch-3-colours.active-colour'

fake_url = 'https://www.strathberry.com/products/east-west-mini-black-with-eyelets'
fake_target_sel = 'oss.swatch-container.swatch-1-colours.active-colour'

target = PageObserver(url=fake_url, target_sel=fake_target_sel)
while True:
    if target.is_changed():
        alert()
        webbrowser.open(fake_url)
        target.driver.close()
    else:
        print('Nope!')

拉勾网职位提醒

其实爬取的时候面向对象还是面向过程， 这个代码可以来对比一下的。
面向过程的时候呢：

面向过程

面向对象可以看做是面向过程的一种扩展，也是一种比较友好的形式。

面向对象

# get_page() -> parse_page() -> filter_job() -> send()
'''
raw_html = []
for i in range(1, 30):
    page = get_page()
    raw_html.append(page)

all_jobs = []
for html in raw_html:
    jobs = parse(html)
    all_jobs.append(jobs)

for job in all_jobs:
    result = filter_job(job)
    if result:
        send(job)
'''

# Spider -> Parser -> Job

'''
s = Spider()
raw_pages = s.crawl(url)
p = Parser(raw_pages)
jobs = p.get_jobs()

for j in jobs:
    if j.is_today():
        j.send_to_me()
'''
from selenium.webdriver import Chrome
from bs4 import BeautifulSoup
import time
# https://www.lagou.com/zhaopin/qukuailian/12/
class Spider:
    def __init__(self, index_url, page_range):
        self.page_range = page_range + 1
        self.index_url = index_url
        self.raw_pages = []
        self.boot()

    def boot(self):
        self.chrome = Chrome(executable_path='./chromedriver')
        self.chrome.start_client()

    def crawl(self):
        for num in range(1, self.page_range):
            full_url = f'{self.index_url}{num}/'
            self.chrome.get(full_url)
            print('Wait for loading page')
            time.sleep(3)
            single_html = self.chrome.page_source
            #本来这个浏览器应该打开就处理，但是也可以一次性网页都加载完再挨个处理，这个可以存内容的。，
            self.raw_pages.append(single_html)
            print('Done')


class Parser:

    def __init__(self, raw_pages):
        self.raw_pages = raw_pages
        self.jobs = []

        self.parse()

    def parse(self):
        for html in self.raw_pages:
            soup = BeautifulSoup(html, 'html.parser') #这里注意是html.parser
            time_sel = 'ul span.format-time'
            comp_sel = 'ul .company_name > a'
            link_sel = 'ul a.position_link'

            time_els = soup.select(time_sel) # list
            comp_els= soup.select(comp_sel)  # list
            link_els= soup.select(link_sel)  # list
            for t,c,l in zip(time_els, comp_els, link_els):

                cell = {  #字典方便查询
                    'time':t.text,
                    'comp':c.text,
                    'link':l.get('href')
                }
                self.jobs.append(cell)
                # [{},{}]
    def get_jobs(self):
        return [Job(j) for j in self.jobs]



class Job:

    def __init__(self,data):

        self.time = data.get('time')
        self.comp = data.get('comp')
        self.link = data.get('link')


    def is_today(self):
        return ':' in self.time # -> T or F

    def send(self):
        pass

    def save_into_csv(self):
        pass


s = Spider(
    index_url='https://www.lagou.com/zhaopin/qukuailian/',
    page_range=2
)
s.crawl()
p = Parser(s.raw_pages)
jobs = p.get_jobs()
for j in jobs:
    if j.is_today():
        print(j.comp,j.link)