[有问题] 使用用户名+密码登录

Login_Bili.py

@@ -0,0 +1,158 @@
+# 有个缺点，我在青龙装不了playwright，所以我就本地弄了个直接写在文件里，然后自动打开文件显示cookie。
+# 第三方包playwright, ddddocr, pillow
+# 安装方式：
+# playwright: 
+#     pip install playwright   # 安装playwright，这个做爬虫比较简单
+#     playwright install       # 安装playwright用到的浏览器内核
+# ddddocr:
+#     pip install ddddocr      # 这个开源识别包挺好用，虽然不是特别准，但是我觉得够用
+# pillow
+#     pip install pillow       # 这个是用来处理图片相关的
+from playwright.sync_api import Playwright, sync_playwright, expect
+import ddddocr
+from PIL import Image
+
+import time
+import os
+from io import BytesIO
+
+################################################################################
+USER_NAME = '用户名'  # 用户名
+USER_PASSWD = '密码'  # 密码
+FILE_ABS_PATH = os.path.abspath(__file__)
+PATH = FILE_ABS_PATH.replace(os.path.split(__file__)[-1], '')  # 我用这个路径存放cookie
+MAX_LOGIN_TRY_TIME = 8
+# 登录按钮位置，这个可以自己使用playwright设置浏览器窗口大小，然后用画图软件查看对应位置
+# 页面 https://passport.bilibili.com/login?gourl=https://space.bilibili.com
+LOGIN_X, LOGIN_Y = 770, 490
+# 验证码点完后点击【确认】位置
+CONFIRM_X, CONFIRM_Y = 970, 550
+# 截取验证码位置
+HEAD_PIC = {"x":880,"y":240,"width":130,"height":40}  # 截取需要点击的文字图片
+CONTENT_X_START, CONTENT_Y_START = 755, 280           # 截取内容图片左上角位置
+CONTENT_PIC = {"x":CONTENT_X_START,"y":CONTENT_Y_START,"width":256,"height":256}  # 截取内容图片
+################################################################################
+class OCR:  # 根据截取的图片，识别要点击的内容，返回要点击的位置
+    def __init__(self) -> None:
+        self.ocr1 = ddddocr.DdddOcr(show_ad=False) # 识别器
+        self.ocr2 = ddddocr.DdddOcr(det=True, show_ad=False) # 目标检测器
+        self.head_ocr = None # 前面文字内容识别结果存储
+        self.content_ocr = dict() # 位置识别结果存储位置
+
+    def __call__(self):
+        self.__ocr_head()  # 识别前面的文字
+        self.__ocr_content()  # 识别有文字的位置
+        return self.__judge()  # 返回要点击的位置
+
+    def __ocr_head(self):  # 文字识别
+        with open(os.path.join(PATH, 'head.png'), "rb")as fp:
+            response = fp.read()
+        self.head_ocr = self.ocr1.classification(response)
+        print(f"HEAD_ddddocr: {self.head_ocr}")
+
+    def __ocr_content(self):  # 内容识别
+        with open(os.path.join(PATH, 'content.png'), "rb")as fp:
+            response = fp.read()
+        img = Image.open(BytesIO(response))
+
+        detect_res = self.ocr2.detection(response)  # 识别文字位置
+        for box in detect_res:
+            x1, y1, x2, y2 = box
+            tmp_img = img.crop((x1, y1, x2, y2))
+            res = self.ocr1.classification(tmp_img)  # 识别文字
+            if len(res):
+                self.content_ocr[res] = ((x1+x2)//2, (y1+y2)//2)
+            print(f"CONTENT_ddddocr: {res} {(x1+x2)//2} {(y1+y2)//2}")
+
+    def __judge(self):
+        '''
+        能识别的且对应的文字确定后，不能识别的就随便给前面识别出来却没用到的位置
+        我在自己电脑试了下，准确率还挺高的，出乎我的预料的好，一般三次内都能成功
+        '''
+        not_exist = []
+        res = []
+        res_content = []
+        for c in range(len(self.head_ocr)):
+            pos = self.content_ocr.get(self.head_ocr[c], None)
+            res.append(pos)
+            if not pos:
+                res_content.append(pos)
+                not_exist.append(c)
+            else:
+                res_content.append(self.head_ocr[c])
+        not_exist_id = 0
+        for d in self.content_ocr.keys():
+            if d not in self.head_ocr:
+                res[not_exist[not_exist_id]] = self.content_ocr[d]
+                res_content[not_exist[not_exist_id]] = d
+                not_exist_id += 1
+                if not_exist_id == len(not_exist):
+                    break
+        return res, res_content
+
+def run(playwright: Playwright) -> None:
+    browser = playwright.chromium.launch(headless=True)  # False会打开浏览器，直接看到操作过程，True不打开浏览器看不到操作过程
+    context = browser.new_context()
+    page = context.new_page()
+    page.goto("https://space.bilibili.com")  # 打开网页
+    try_time = 0  # 设置登录尝试次数
+    while True:
+        try_time += 1
+        if try_time == MAX_LOGIN_TRY_TIME:
+            print("登录尝试次数过多，退尝试！")
+            exit()
+        print(f'{"-"*42}第[{try_time}]次登录尝试{"-"*42}')
+        # 输入账号密码
+        page.get_by_placeholder("你的手机号/邮箱").fill(USER_NAME)
+        page.get_by_placeholder("密码").fill(USER_PASSWD)
+        time.sleep(1)
+        # 模拟鼠标点击登录按钮
+        page.mouse.click(LOGIN_X, LOGIN_Y)
+        time.sleep(3)
+        print(page.title())
+        # 截屏
+        page.screenshot(full_page=True, path=os.path.join(PATH, 'B.png'))  # 截取整个页面，保存为B.png
+        page.screenshot(path=os.path.join(PATH, 'head.png'), clip=HEAD_PIC)  # 截取上面的文字顺序，保存为head.png
+        page.screenshot(path=os.path.join(PATH, 'content.png'), clip=CONTENT_PIC)  # 截取下面要点击的区域，保存为content.png
+        # 识别
+        ocr = OCR()
+        pos, content = ocr()  # 返回要点击的位置与对应的内容（测试的时候看）
+        print(pos, content)
+        for p in pos:
+            if p:
+                # 模拟鼠标对应位置
+                page.mouse.click(p[0]+CONTENT_X_START, p[1]+CONTENT_Y_START)
+                time.sleep(3)
+        page.mouse.click(CONFIRM_X, CONFIRM_Y)
+        time.sleep(3)
+        page.goto("https://space.bilibili.com")
+        if '哔哩哔哩弹幕视频网' not in page.title():  # 判断是否成功登录
+            print(page.title())
+            break
+        print('=' * 100)
+    # 读取cookie相关信息
+    required_cookie = ['SESSDATA', 'bili_jct', 'DedeUserID', 'DedeUserID__ckMd5', 'sid']
+    cookie_lst = []
+    for x in context.cookies():
+        name = x.get('name', None)
+        value = x.get('value', None)
+        if name in required_cookie:
+            cookie_lst.append(f'{name}={value}')
+            required_cookie.remove(name)
+    cookie_str = ';'.join(cookie_lst)
+    print('Cookie:')
+    print(cookie_str)
+    # 将cookie放在文件cookie.txt中，默认路径和本文件路径相同
+    with open(os.path.join(PATH, 'cookie.txt'), 'w') as f:
+        f.write(cookie_str)
+    os.popen(f'code {os.path.join(PATH, "cookie.txt")}')
+    # 关闭爬虫浏览器
+    context.close()
+    browser.close()
+
+with sync_playwright() as playwright:
+    run(playwright)
+    # 删除中间截图的图片
+    os.remove(os.path.join(PATH, "B.png"))
+    os.remove(os.path.join(PATH, "content.png"))
+    os.remove(os.path.join(PATH, "head.png"))

qinglong/extra.sh

@@ -5,6 +5,4 @@
 # 安装 dotnet 环境
 # curl -sSL https://ghproxy.com/https://raw.githubusercontent.com/RayWangQvQ/BiliBiliToolPro/main/qinglong/ray-dotnet-install.sh | bash /dev/stdin --no-official
 curl -sSL https://ghproxy.com/https://raw.githubusercontent.com/RayWangQvQ/BiliBiliToolPro/main/qinglong/ray-dotnet-install.sh | bash /dev/stdin
-# 安装需要用到的包
-pip3 install requests qrcode pillow
 # 其他代码...

src/Ray.BiliBiliTool.Application/LoginTaskAppService.cs

@@ -233,7 +233,10 @@ protected void AddOrUpdateCkToQingLong(BiliCookie ckInfo)
             }
 
             _logger.LogInformation("用户不存在，新增cookie");
-            var lastNum = list.LastOrDefault()?.value.Split("__").LastOrDefault();
+            var lastNum = list.LastOrDefault()
+                ?.name
+                .Split("__")
+                .LastOrDefault();
             var newNum = int.Parse(lastNum ?? "-1") + 1;
             var name = $"Ray_BiliBiliCookies__{newNum}";