初始化导入

admin

2025-03-31 4603536999de113969dd7164f3a04dcde9083480

 comment.py

@@ -1,4 +1,3 @@
import re
import threading
import time

@@ -9,7 +8,9 @@
from selenium.webdriver.support.wait import WebDriverWait
from selenium.webdriver.support import expected_conditions as EC

import comment_manager
import setting
from video_manager import VideoManger


class CommentManager:
@@ -50,20 +51,19 @@
        # 等到左侧菜单出现后才能执行后续操作
        wait = WebDriverWait(self.driver, 100)  # 最多等待10秒
        element = wait.until(EC.visibility_of_element_located((By.ID, "side-bar")))
        threading.Thread(target=lambda: self.click_like(self.driver), daemon=True).start()
        threading.Thread(target=lambda: self.process_videos(self.driver), daemon=True).start()

    def __process_like(self, comment_content, comment_element):
    def __process_like(self, video_name, video_date, comment_element):
        """
        处理单条评论的赞
        :param comment_content:
        :param comment_element:
        :return:
        """
        comment_pattern = re.compile(r'[\u4e00-\u9fa5]+')
        comment_content = re.sub("<img.*?>", "", comment_content)
        if not comment_pattern.search(comment_content):
            # 符合标准的评论
        user_name, comment_content, comment_time = self.__parse_comment_info(comment_element)
        if not comment_manager.is_need_click_like(video_name, video_date, user_name, comment_content):
            return

        comment_actions = comment_element.find_element(By.CLASS_NAME, "action-list").find_elements(By.CLASS_NAME,
                                                                                                   "action-item")
        classes: str = comment_actions[0].find_element(By.TAG_NAME, "svg").get_attribute("class")
@@ -72,50 +72,73 @@
            return
        like = comment_element.find_element(By.CLASS_NAME, "like-action")
        like.click()
        self.driver.implicitly_wait(2)
        # self.driver.implicitly_wait(2)
        time.sleep(2)

    def __process_reply(self, comment_content, comment_element):
    def __parse_comment_info(self, comment_element):
        """
        处理单条评论的回复
        :param comment_content:
        解析评论信息
        :param comment_element:
        :return:
        :return:(评论人,评论内容,评论时间)
        """
        user_name = comment_element.find_element(By.CLASS_NAME, "comment-user-name").get_attribute("innerHTML")
        comment_time = comment_element.find_element(By.CLASS_NAME, "comment-time").get_attribute("innerHTML")
        comment_content = comment_element.find_element(By.CLASS_NAME, "comment-content").get_attribute("innerHTML")
        return user_name, comment_content, comment_time

    def __parse_video_info(self, video_element):
        """
        解析视频信息
        :param comment_element:
        :return:(视频名称,视频时间,评论数量)
        """
        video_time = video_element.find_element(By.CLASS_NAME, "feed-time").get_attribute("innerHTML")
        comment_count = video_element.find_element(By.CLASS_NAME, "feed-comment-total").find_element(By.XPATH,
                                                                                                     "./span[2]").get_attribute(
            "innerHTML")
        video_name = video_element.find_element(By.CLASS_NAME, "comment-content").get_attribute("innerHTML")
        return video_name, video_time, comment_count

    def __process_reply(self, video_name, video_date, comment_element):
        """
        处理单条评论回复
        :param video_name:
        :param video_date:
        :param comment_element:
        :return: 是否需要点赞
        """

        def get_reply_comment(nick_name, content):
            if not setting.is_reply_comment():
                return None
            # 内容是否符合标准
            for t in self.comment_templates:
                if re.match(t[0], content):
                    # 满足内容
                    return t[1].replace("[昵称]", nick_name)
            return comment_manager.get_replay_content(video_name, video_date, nick_name, content)

        comment_pattern = re.compile(r'[\u4e00-\u9fa5]+')
        comment_content = re.sub("<img.*?>", "", comment_content)
        if not comment_pattern.search(comment_content):
            # 符合标准的评论
            return
        comment_user_name, comment_content, comment_time = self.__parse_comment_info(comment_element)

        try:
            comment_reply_list = comment_element.find_element(By.CLASS_NAME, "comment-reply-list")
            if comment_reply_list.find_element(By.XPATH, ".//div[normalize-space()='作者']"):
                print("已经回复。。。。")
                return
                return False
        except:
            print("还没回复。。。。")

        replay_content = get_reply_comment(comment_user_name, comment_element)
        if not replay_content:
            # 不需要评论
            print("不需要回复。。。。")
            return True

        # 需要回复

        comment_actions = comment_element.find_element(By.CLASS_NAME, "action-list").find_elements(By.CLASS_NAME,
                                                                                                   "action-item")
        # 点击评论
        comment_actions[1].click()
        wait = WebDriverWait(self.driver, 5)  # 最多等待10秒
        element = wait.until(EC.visibility_of_element_located((By.CLASS_NAME, "comment-create-content")))
        # TODO 昵称
        replay_content = get_reply_comment("昵称", comment_element)
        if not replay_content:
            # 不需要评论
            return

        wait.until(EC.visibility_of_element_located((By.CLASS_NAME, "comment-create-content")))
        time.sleep(1)
        self.driver.find_element(By.CLASS_NAME,
                                 "comment-create-content").find_element(By.TAG_NAME,
                                                                        "textarea").send_keys(replay_content)
@@ -123,8 +146,16 @@
        self.driver.find_element(By.CLASS_NAME,
                                 "comment-create-content").find_element(By.XPATH, "div[3]/div[2]").click()
        time.sleep(2)
        return False

    def __click_like_all(self, driver, start_index=0):
    def __process_comments(self, driver, start_index=0, video_info=None):
        """

        :param driver:
        :param start_index:
        :param video_info:(视频内容, 视频日期, 评论次数)
        :return:
        """
        scroll_list = driver.find_element(By.CLASS_NAME, "feed-comment__wrp")

        loadmore = scroll_list.find_element(By.CLASS_NAME, "loadmore__dot")
@@ -137,34 +168,60 @@
        for index in range(start_index, len(comments)):
            comment = comments[index]
            # 评论内容
            comment_content = comment.find_element(By.CLASS_NAME, "comment-content").get_attribute("innerHTML")
            user_name, comment_content, comment_time = self.__parse_comment_info(comment)
            if not comment_content:
                continue
            print("=======评论内容：", comment_content)
            self.__process_reply(comment_content, comment)
            self.__process_like(comment_content, comment)
            need_like = self.__process_reply(video_info[0], video_info[1], comment)
            if need_like:
                print("需要点赞")
                self.__process_like(video_info[0], video_info[1], comment)
            else:
                print("不需要点赞")
        # 往下滑动
        if has_more:
            driver.execute_script(
                "var __comment_scroll = document.getElementsByClassName('feed-comment__wrp')[0]; __comment_scroll.scrollTop = __comment_scroll.scrollHeight;")
            print("往下滚动")
            time.sleep(3)
            self.__click_like_all(driver, start_index=len(comments))
            self.__process_comments(driver, start_index=len(comments))
        else:
            print("没有更多了")
            print("没有更多评论了")

    def click_like(self, driver: webdriver.Chrome):
    def process_videos(self, driver: webdriver.Chrome, start_index=0):
        wait = WebDriverWait(driver, 100)  # 最多等待10秒
        element = wait.until(EC.visibility_of_element_located((By.CLASS_NAME, "feeds-container")))
        scroll_list = wait.until(EC.visibility_of_element_located((By.CLASS_NAME, "feeds-container")))
        time.sleep(1)
        # 查找视频列表
        videos_root = driver.find_element(By.CLASS_NAME, "feeds-container")
        videos = videos_root.find_elements(By.CLASS_NAME, "comment-feed-wrap")
        for video in videos:
        for i in range(start_index, len(videos)):
            # 选择视频
            video = videos[i]
            # 解析视频内容
            video_name, video_time, comment_count = self.__parse_video_info(video)
            # 判断是否要点击进去
            if not VideoManger().is_need_click(video_name, video_time, comment_count):
                continue
            video.click()
            driver.implicitly_wait(2)
            self.__click_like_all(driver)
            self.__process_comments(driver, video_info=(video_name, video_time, comment_count))
            # 5s处理一个视频
            time.sleep(5)

        loadmore = scroll_list.find_element(By.CLASS_NAME, "loadmore__dot")
        # 获取父控件
        loadmore = loadmore.find_element(By.XPATH, "./..")
        has_more = loadmore.value_of_css_property("display") == "none"
        if has_more:
            driver.execute_script(
                "var __video_scroll = document.getElementsByClassName('feeds-container')[0]; __video_scroll.scrollTop = __video_scroll.scrollHeight;")
            print("往下滚动")
            time.sleep(3)
            self.process_videos(driver, start_index=len(videos))
        else:
            print("没有更多视频了")
            VideoManger().add_video_infos([self.__parse_video_info(v) for v in videos])

    def close(self):
        if self.driver: