不会玩阴阳师的我带你一键下载《阴阳师：百闻牌》所有卡牌并调用百度OCR识别文字信息

import time
from selenium import webdriver
from selenium.webdriver import ActionChains
from selenium.webdriver.common.keys import Keys

driver = webdriver.Chrome()
driver.maximize_window()
driver.get('https://ssr.163.com/cardmaker/#/')
time.sleep(3)
while True:
    try:
        driver.find_element_by_class_name('load-more')
        ActionChains(driver).key_down(Keys.END).perform()
        time.sleep(1)
    except:
        break
# 通过xpath定位所有图片
imgs = driver.find_elements_by_xpath('/html/body/div[3]/div[4]/div/div[3]/ul/li/img')
print('共计有%d张卡牌' % len(imgs))
for img in imgs:
    print('已获取到图片链接：', img.get_attribute('src'))
driver.quit()

next_page = driver.find_element_by_class_name('//*[@id="mainsrp-pager"]/div/div/div/ul/li[8]/a/span[1]')
        next_page.click()

url = 'https://s.taobao.com/search?q=Python&imgfile=&commend=all&ssid=s5-e&search_type=item&sourceId=tb.index&spm=a21bo.1000386.201856-taobao-item.1&ie=utf8&initiative_id=tbindexz_20170306&bcoffset=0&ntoffset=6&p4ppushleft=1%2C48&s={}'.format(44 * (i - 1)) # i为页数

while True:
	try:
    	driver.find_element_by_xpath('load-more').click()
   	except:
   		break

js = "var q=document.documentElement.scrollTop=100000"
while True:
    try:
        driver.find_element_by_class_name('load-more')
        driver.execute_script(js)
        time.sleep(1)
    except:
        break

while True:
    try:
        driver.find_element_by_class_name('load-more')
        ActionChains(driver).key_down(Keys.END).perform()
        time.sleep(1)
    except:
        break

import csv
import requests
import time
import os
from selenium import webdriver
from selenium.webdriver.common.action_chains import ActionChains
from selenium.webdriver.common.keys import Keys
from concurrent.futures import ThreadPoolExecutor
from aip import AipOcr

# 定义失败数量对文字识别失败进行计数
fail_num = 0
# 定义百度api参数
APP_ID = '17076767'
API_KEY = 'Af3Rj5HALMz5AN8prSgwTH4m'
SECRET_KEY = '49lYvSCocGikjWTbYFuFuGzK9Eph8xqv'
# 定义文字识别参数
options = {
    'detect_direction': 'true',
    'language_type': 'CHN_ENG',
}
# 初始化AipOcr对象
aipOcr = AipOcr(APP_ID, API_KEY, SECRET_KEY)
# 定义请求头
headers = {
    'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/80.0.3987.116 Safari/537.36'
}

def slide_down_key(driver):
    '''模拟按键模拟滚动到页面底部'''
    while True:
        try:
            driver.find_element_by_class_name('load-more')
            ActionChains(driver).key_down(Keys.END).perform()
            time.sleep(1)
        except:
            break


def slide_down_js(driver):
    '''执行JS模拟滚动到页面底部'''
    js = "var q=document.documentElement.scrollTop=100000"
    while True:
        try:
            driver.find_element_by_class_name('load-more')
            driver.execute_script(js)
            time.sleep(1)
        except:
            break


def get_img_src():
    '''在页面中获取所有卡牌链接'''
    driver = webdriver.Chrome()
    driver.maximize_window()
    driver.get('https://ssr.163.com/cardmaker/#/')
    time.sleep(3)
    slide_down_key(driver)
    # slide_down_js(driver)
    # 通过xpath定位所有图片
    imgs = driver.find_elements_by_xpath('/html/body/div[3]/div[4]/div/div[3]/ul/li/img')
    print('共计有%d张卡牌' % len(imgs))
    for img in imgs:
        print('已获取到图片链接：', img.get_attribute('src'))
        yield img.get_attribute('src')
    driver.quit()

def download_pic(index, url, writer):
    '''下载图片'''
    content = requests.get(url).content
    file_name = url.split('/')[-1].split('.')[0]
    print('当前正在下载第%d张图片：%s' % (index, file_name))
    with open('Cards/%s.png' % file_name, 'wb') as f:
        f.write(content)
    time.sleep(0.5)
    return file_name, index, writer


def identify_text(result):
    '''调用百度aip识别文字并保存'''
    global fail_num
    time.sleep(0.5)
    file_name = result.result()[0]
    index = result.result()[1]
    writer = result.result()[2]
    print('当前正在识别第%d张图片：%s' % (index, file_name))
    try:
        with open('Cards/%s.png' % file_name, 'rb') as fp:
            # 读取图片内容并用aip识别获取返回内容
            pic_content = fp.read()
            result = aipOcr.general(pic_content, options)
            # 从结果中得到文字、对应图片区域的顶部位置和高度
            words_result = [i['words'] for i in result['words_result']]
            top_locations = [int(i['location']['top']) for i in result['words_result']]
            height_locations = [int(i['location']['height']) for i in result['words_result']]
            result_length = len(words_result)
            # 初始化角色、技能和卡牌描述
            if len(words_result[0]) >= 2:
                role = words_result[0]
            else:
                role = words_result[1]
            if len(words_result[-1]) >= 2:
                skill = words_result[-1]
            else:
                skill = words_result[-2]
            card_desp = ''
            for i in range(result_length):
                # 文字区域大于45为数字，过滤
                if height_locations[i] >= 45:
                    continue
                # 顶部位置小于340为无文字区域，跳过
                if top_locations[i] <= 340:
                    continue
                # 根据顶部和高度获取角色
                if (top_locations[i] + height_locations[i]) <= 395:
                    role = words_result[i]
                    continue
                # 根据顶部位置获取技能
                if top_locations[i] >= 515:
                    skill = words_result[i]
                    continue
                # 获取卡牌描述字符串并拼接
                card_desp += words_result[i]
            print('当前角色是：', role)
            writer.writerow([role, card_desp, skill])
    # 异常处理
    except Exception as e:
        print(e.args[0])
        fail_num += 1

def main():
    '''主函数'''
    # 定义线程池，实现多线程
    pool = ThreadPoolExecutor(2)
    # 创建文件夹
    if not os.path.exists('Cards'):
        os.mkdir('Cards')
    if os.path.exists('Card_Data.csv'):
        os.remove('Card_Data.csv')
    # 创建csv文件用于保存数据
    with open('Card_Data.csv', 'a+', newline='', encoding='gb18030') as f:
        writer = csv.writer(f, dialect="excel")
        writer.writerow(['角色', '描述', '技能'])
        for index, img_url in enumerate(get_img_src()):
            # 向线程池中添加下载任务并在任务完成后回调实现文字识别
            pool.submit(download_pic, index + 1, img_url, writer).add_done_callback(identify_text)
        pool.shutdown(wait=True)
    print('未成功识别文字的图片为：', fail_num)


if __name__ == '__main__':
    '''执行主函数'''
    print('******程序开始运行******')
    start = time.time()
    main()
    time = int(time.time() - start)
    print('******程序执行结束，共计用时%d分%d秒。******' % (time // 60, time % 60))