from selenium import webdriver
from selenium.webdriver.common.action_chains import ActionChains
import requests
import time
import numpy
import cv2
import os

#作者：許文鴻
#未經(jīng)允許不可轉(zhuǎn)載，轉(zhuǎn)載時(shí)注明出處

#創(chuàng)建 WebDriver 對(duì)象，指明使用chrome瀏覽器驅(qū)動(dòng)
web = webdriver.Chrome(r'd:\chromedriver.exe')
web.implicitly_wait(5)
#調(diào)用WebDriver 對(duì)象的get方法 可以讓瀏覽器打開指定網(wǎng)址
web.get('https://pintia.cn/auth/login')
zh = web.find_element_by_xpath('/html/body/div[1]/div[3]/div/div[2]/form/div[1]/div[1]/div/div/div[1]/input')
mm = web.find_element_by_xpath('/html/body/div[1]/div[3]/div/div[2]/form/div[1]/div[2]/div/div/div[1]/input')

#在PTA的賬號(hào)密碼：
zh.send_keys('******@qq.com')
mm.send_keys('******')
#找到登錄按鈕并點(diǎn)擊
web.find_element_by_xpath('/html/body/div[1]/div[3]/div/div[2]/form/div[2]/button/div/div').click()
#等待兩秒，驗(yàn)證碼加載完成
time.sleep(2)
#bg背景圖片
bg_img_src = web.find_element_by_xpath(
 '/html/body/div[3]/div[2]/div/div/div[2]/div/div[1]/div/div[1]/img[1]').get_attribute('src')
#front可拖動(dòng)圖片
front_img_src = web.find_element_by_xpath(
 '/html/body/div[3]/div[2]/div/div/div[2]/div/div[1]/div/div[1]/img[2]').get_attribute('src')
#保存圖片
with open("bg.jpg", mode="wb") as f:
 f.write(requests.get(bg_img_src).content)
with open("front.jpg", mode="wb") as f:
 f.write(requests.get(front_img_src).content)
#將圖片加載至內(nèi)存
bg = cv2.imread("bg.jpg")
front = cv2.imread("front.jpg")
js = 'alert("本人可能將此程序用于python課設(shè)，請(qǐng)靚仔靚女不要直接提交本人代碼。即將報(bào)錯(cuò)，需要?jiǎng)h除第42~44行代碼即可正常運(yùn)行");'
web.execute_script(js)
time.sleep(15)
#將背景圖片轉(zhuǎn)化為灰度圖片，將三原色降維
bg = cv2.cvtColor(bg, cv2.COLOR_BGR2GRAY)
#將可滑動(dòng)圖片轉(zhuǎn)化為灰度圖片，將三原色降維
front = cv2.cvtColor(front, cv2.COLOR_BGR2GRAY)
front = front[front.any(1)]
#用cv算法匹配精度最高的xy值
result = cv2.matchTemplate(bg, front, cv2.TM_CCOEFF_NORMED)
#numpy解析xy，注意xy與實(shí)際為相反，x=y,y=x
x, y = numpy.unravel_index(numpy.argmax(result), result.shape)
#找到可拖動(dòng)區(qū)域
div = web.find_element_by_xpath('/html/body/div[3]/div[2]/div/div/div[2]/div/div[2]/div[2]')
#拖動(dòng)滑塊，以實(shí)際相反的y值代替x
ActionChains(web).drag_and_drop_by_offset(div, xoffset=y // 0.946, yoffset=0).perform()

#至此成功破解驗(yàn)證碼，由于算法問題，準(zhǔn)確率不能達(dá)到100%，可能需要多運(yùn)行1~2次

for page in range(0, 1000):
 time.sleep(1)
 #此處的網(wǎng)址為PTA固定網(wǎng)頁，僅需要更換page
 web.get('https://pintia.cn/problem-sets?tab=1&filter=all&page={page_}'.format(page_=page))
 #獲取當(dāng)前頁面題目集網(wǎng)址，A_s為a標(biāo)簽的列表，urls用戶存放網(wǎng)址
 A_s = web.find_elements_by_class_name('name_QIjv7')
 urls = []
 for a in A_s:
  urls.append(a.get_attribute('href'))
 #當(dāng)頁面不存在可爬取的網(wǎng)址，則退出程序
 if urls.__len__() == 0:
  print('爬取完成')
  os._exit()
 #對(duì)剛才獲取的網(wǎng)址列表進(jìn)行遍歷
 for url in urls:
  web.get(url)
  #找到對(duì)應(yīng)的題目對(duì)象
  tm = web.find_elements_by_css_selector("[class='problemStatusRect_3kpmC PROBLEM_ACCEPTED_1Dzzi']")
  tm_total = 0
  for i in range(0, 1000):
   # 遍歷該頁面的題型
   try:
    tm_type = web.find_element_by_xpath(
     '/html/body/div/div[3]/div[2]/div/div[2]/div[{i_}]/div/div[2]'.format(i_=i * 2 + 2)).text
    # 如果題型為編程/函數(shù)，記錄對(duì)應(yīng)的數(shù)量，方便后續(xù)爬取
    if tm_type == '編程題' or tm_type == '函數(shù)題':
     tm_total += int(web.find_element_by_xpath(
      '/html/body/div/div[3]/div[2]/div/div[2]/div[{i_}]/a/div/div'.format(i_=i * 2 + 2)).text[0])
   except:
    break
  # 根據(jù)函數(shù)/編程題數(shù)量取相應(yīng)的題目對(duì)象，舍棄其他題目
  if tm_total != 0:
   tm = tm[-tm_total:]
  else:
   tm = []
  # 遍歷剩余題目
  for tm_index in tm:
   try:
    tm_index.click()
    time.sleep(0.5)
    #獲取題目中的代碼
    tm_title = web.find_element_by_css_selector(
     "[class='text-center black-3 text-4 font-weight-bold my-3']").text
    mycode = web.find_element_by_css_selector('textarea').get_attribute('value')
    print('題目：' + tm_title)
    print(mycode)
    #接下來可以存入
   except:
    continue