1.购买服务器阿里云：服务器购买地址https://t.aliyun.com/U/W6Zbs4若失效，可用地址
阿里云：
服务器购买地址
https://t.aliyun.com/U/W6Zbs4若失效，可用地址
https://www.aliyun.com/minisite/goods?source=5176.29345612&userCode=49hts92d腾讯云：
https://curl.qcloud.com/wJpWmSfU若失效，可用地址
https://cloud.tencent.com/act/cps/redirect?redirect=2446&cps_key=ad201ee2ef3b771157f72ee5464b1fea&from=console华为云
https://activity.huaweicloud.com/cps.html?fromacct=64b5cf7cc11b4840bb4ed2ea0b2f4468&utm_source=V1g3MDY4NTY=&utm_medium=cps&utm_campaign=2019052.部署教程
3.代码如下
import timefrom selenium import webdriverfrom selenium.webdriver.chrome.service import Servicefrom selenium.webdriver.chrome.options import Optionsfrom selenium.webdriver.common.by import Byfrom selenium.webdriver.support.wait import WebDriverWaitimport pdfplumberimport reimport osfrom openai import OpenAI# 全局变量count = 0# 读取键盘输入并赋值给intervaldef get_user_input():print("请选择模式：")print("1. 测试")print("2. 考试")mode = int(input("请输入模式编号（1或2）："))print("请输入做题间隔时间(单位为秒)：")interval = float(input())print("请输入做题总数：")total = int(input())print(f"本次时间间隔为：{interval}s，预期答题时间：{interval * total + 60}s")return mode, interval, total# 设置DataBasedef extract_text_from_pdf(pdf_path, start_page, end_page):text_list = []with pdfplumber.open(pdf_path) as pdf:for page_number, page in enumerate(pdf.pages, start=1):if end_page is not None and page_number > end_page:breakif page_number >= start_page:text = page.extract_text()if text:text_list.append(text)return text_listdef parse_text_to_dict(text_list):word_dict = {}pattern = re.compile(r'(\b\w+\b)\s+(.*)')for text in text_list:lines = text.splitlines()for line in lines:match = pattern.match(line)if match:word = match.group(1)meaning = match.group(2).strip()word_dict[word] = meaningreturn word_dictdef parse_text_to_dict_NI(text_list):word_dict = {}pattern = re.compile(r'(\b\w+\b)\s+(.*)')for text in text_list:lines = text.splitlines()for line in lines:match = pattern.match(line)if match:meaning = match.group(2)word = match.group(1).strip()word_dict[word] = meaningreturn word_dict# 初始化浏览器def init_browser():mobile_emulation = {"deviceMetrics": {"width": 1707,"height": 773,"pixelRatio": 1.0},"userAgent": "Mozilla/5.0 (Linux; Android 13; IQOO 18) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/116.0.0.0 Mobile Safari/537.36"}chrome_options = Options()chrome_options.add_experimental_option("mobileEmulation", mobile_emulation)wd = webdriver.Chrome(service=Service(r"./chromedriver.exe"), options=chrome_options)return wd# 开始测试或考试def start_test_or_exam(wd, interval, total, mode):global countfor i in range(total):question = WebDriverWait(wd, 15, 0.5).until(lambda wd: wd.find_element(By.XPATH, '//*[@id="app"]/div/div[3]/div/div[2]/div/div[1]/span[2]'))question = question.text.split()[0].strip()A = WebDriverWait(wd, 15, 0.1).until(lambda wd: wd.find_element(By.XPATH, '//*[@id="app"]/div/div[3]/div/div[3]/div/div[1]/div[1]/span'))B = WebDriverWait(wd, 15, 0.1).until(lambda wd: wd.find_element(By.XPATH, '//*[@id="app"]/div/div[3]/div/div[3]/div/div[2]/div[1]/span'))C = WebDriverWait(wd, 15, 0.1).until(lambda wd: wd.find_element(By.XPATH, '//*[@id="app"]/div/div[3]/div/div[3]/div/div[3]/div[1]/span'))D = WebDriverWait(wd, 15, 0.1).until(lambda wd: wd.find_element(By.XPATH, '//*[@id="app"]/div/div[3]/div/div[3]/div/div[4]/div[1]/span'))At = A.text.split('.')[1].strip()Bt = B.text.split('.')[1].strip()Ct = C.text.split('.')[1].strip()Dt = D.text.split('.')[1].strip()con = f"""{question}\nA.{At}\nB.{Bt}\nC.{Ct}\nD.{Dt}"""print(con)if question.isalpha():flag = 1for word, meaning in word_dict.items():if question == word:if At in meaning and Bt not in meaning and Ct not in meaning and Dt not in meaning:flag = 0A.click()breakif Bt in meaning and At not in meaning and Ct not in meaning and Dt not in meaning:flag = 0B.click()breakif Ct in meaning and Bt not in meaning and At not in meaning and Dt not in meaning:flag = 0C.click()breakif Dt in meaning and Bt not in meaning and Ct not in meaning and At not in meaning:flag = 0D.click()breakif flag:count += 1client = OpenAI(api_key=os.getenv("DEEPSEEK_API_KEY"),base_url="https://api.deepseek.com")completion = client.chat.completions.create(model="deepseek-chat",messages=[{'role': 'system', 'content': '你要做的是词义匹配，找到和英文单词最贴切的中文解释，最终回答一个用"-"包起来的大写字母作为答案,例如"-B-"'},{'role': 'user', 'content': con},])WebDriverWait(wd, 10, 0.1).until(lambda wd: '-' in completion.choices[0].message.content)print(f"-----(英译中){i + 1} " + completion.choices[0].message.content)if '-A-' in completion.choices[0].message.content:A.click()if '-B-' in completion.choices[0].message.content:B.click()if '-C-' in completion.choices[0].message.content:C.click()if '-D-' in completion.choices[0].message.content:D.click()completion.choices[0].message.content = Noneelse:flag = 1for word, meaning in word_dict_NI.items():if question in word:if At in meaning:flag = 0A.click()breakif Bt in meaning:flag = 0B.click()breakif Ct in meaning:flag = 0C.click()breakif Dt in meaning:flag = 0D.click()breakif flag:count += 1client = OpenAI(api_key=os.getenv("DEEPSEEK_API_KEY"),base_url="https://api.deepseek.com")completion = client.chat.completions.create(model="deepseek-chat",messages=[{'role': 'system', 'content': '你要做的是词义匹配，找到和中文意思最贴切的英语单词，最终回答一个用"-"包起来的大写字母作为答案,例如"-B-"'},{'role': 'user', 'content': con},])WebDriverWait(wd, 10, 0.1).until(lambda wd: '-' in completion.choices[0].message.content)print(f"-----(中译英){i + 1} " + completion.choices[0].message.content)if '-A-' in completion.choices[0].message.content:A.click()if '-B-' in completion.choices[0].message.content:B.click()if '-C-' in completion.choices[0].message.content:C.click()if '-D-' in completion.choices[0].message.content:D.click()completion.choices[0].message.content = Nonetime.sleep(interval)time.sleep(0.1)if mode == 2: # 如果是考试模式，提交答案wd.find_element(By.XPATH, '//*[@id="app"]/div/div[2]/div/div/div[3]/span').click()time.sleep(0.1)wd.find_element(By.XPATH, '/html/body/div[4]/div[2]/button[2]').click()print(f"使用了{count}次AI")time.sleep(999999)# 主函数def main():mode, interval, total = get_user_input()print("Database Setting>>>.....")pdf_path = r'./Data.pdf'start_page = 1end_page = 113text_list = extract_text_from_pdf(pdf_path, start_page, end_page)global word_dict, word_dict_NIword_dict = parse_text_to_dict(text_list)word_dict_NI = parse_text_to_dict_NI(text_list)for word, meaning in word_dict.items():print(f"{word}=={meaning}")for word, meaning in word_dict_NI.items():print(f"{word}=={meaning}")print("Task Beginning>>>.....")wd = init_browser()wd.get("https://skl.hduhelp.com/?type=5#/english/list")wd.maximize_window()time.sleep(2.5)began = WebDriverWait(wd, 150, 0.1).until(lambda wd: wd.find_element(By.XPATH, '//*[@id="app"]/div/div[3]/div/div[2]/div[2]/div/button'))began.click()start_test_or_exam(wd, interval, total, mode)if __name__ == "__main__":main()
解析
该脚本为我爱记歌词自动答题脚本，脚本主要作用：
通过 Selenium 打开指定网站的英语题库页面，自动开始做题（支持"测试/考试"两种模式）。
先从本地 PDF（Data.pdf） 提取"单词-释义"数据，建立词典用于本地匹配答案。
若本地词典无法唯一确定答案，则调用 DeepSeek Chat API（通过
openai客户端、DEEPSEEK_API_KEY环境变量）让模型判定并选项自动点击。根据设定的做题间隔与做题总数循环答题；考试模式结束后自动提交。
记录并输出本次使用 AI 判题的次数。
主要函数/方法作用
get_user_input()
交互式读取模式（1测试/2考试）、两题间隔秒数、总题数，并打印预计用时。extract_text_from_pdf(pdf_path, start_page, end_page)
用 pdfplumber 读取 PDF 指定页区间的文本，返回按页组成的文本列表。parse_text_to_dict(text_list)
将 PDF 文本解析为英→中词典：以每行"英文 单行中文释义"格式，构建{英文: 中文释义}。parse_text_to_dict_NI(text_list)
与上类似，但用于中→英方向匹配：构建{英文: 中文释义}，供中文题干反查英文选项。init_browser()
配置移动端 UA 与分辨率，启动chromedriver.exe并返回webdriver.Chrome实例。start_test_or_exam(wd, interval, total, mode)
核心答题循环：抓取题干与 A/B/C/D 选项；
若题干为纯字母（英译中），用
word_dict进行本地匹配；无法唯一确定时调用 DeepSeek 让模型返回-A-/ -B- / -C- / -D-，再自动点击；否则视为中文题干（中译英），在
word_dict_NI中查找包含关系；不确定时同样调用 DeepSeek 判定；每题间
time.sleep(interval)；若模式为考试（
mode==2），结束后自动点击交卷；打印本次使用 AI 的次数。
main()
调用get_user_input()读取参数 → 解析 PDF 生成两套词典 → 启动浏览器 → 打开目标页面并点击"开始" → 调用start_test_or_exam()执行全流程。
备注：脚本依赖
selenium,pdfplumber,openai（DeepSeek 兼容接口）,chromedriver.exe。需要在环境中设置DEEPSEEK_API_KEY；并确保Data.pdf存在且页码范围正确。
注意：
本文部分变量已做脱敏处理，仅用于测试和学习研究，禁止用于商业用途，不能保证其合法性，准确性，完整性和有效性，请根据情况自行判断。技术层面需要提供帮助，可以通过打赏的方式进行探讨。
没有评论:
发表评论