craw_spider/captcha/captcha_deal.py

import base64
import random
import time
from urllib.parse import quote

import requests
from DrissionPage import ChromiumPage, ChromiumOptions
from PIL import Image

token = "zPzmt1mG1ouCU6GTzsZN2Lmm8pdZypapPcLJTBRETco"

proxyUrl = "117.26.226.31:40030"
proxies = {
    "http": "http://" + proxyUrl,
    "https": "http://" + proxyUrl,
}

chrome_path = r"C:\Program Files\Google\Chrome\Application\chrome.exe"

class CaptchaDeal(object):
    def __init__(self):
        self.page = self.init_chromepage()

    def init_chromepage(self):
        # 初始化ChromiumPage
        co = ChromiumOptions().auto_port().set_browser_path(chrome_path)
        co.headless(False)  # 可视化模式便于调试

        # co.set_proxy("http://" + proxyUrl)
        page = ChromiumPage(co)
        return page

    def get_page(self):
        """解析验证码界面"""
        url = "https://cfe.m.jd.com/privatedomain/risk_handler/03101900/?returnurl=https%3A%2F%2Fsearch.jd.com%2FSearch%3Fkeyword%3D%25E5%25A4%25A9%25E5%25A3%25AB%25E5%258A%259B%25E5%2585%25BB%25E8%25A1%2580%25E6%25B8%2585%25E8%2584%2591%25E9%25A2%2597%25E7%25B2%25924g%26enc%3Dutf-8%26wq%3D%25E5%25A4%25A9%25E5%25A3%25AB%25E5%258A%259B%25E5%2585%25BB%25E8%25A1%2580%25E6%25B8%2585%25E8%2584%2591%25E9%25A2%2597%25E7%25B2%25924g&rqhost=https%3A%2F%2Fapi.m.jd.com&rpid=rp-112038530-10428-1774874729116&evtype=2&evapi=color_pc_search_searchWare&source=1&forceCurrentView=1&evsid=AASWz-Ccn-fmDu0QbI5tPvKhPINYZNJcz6vQ-TCMSXxv5jRuoOMT4SCV_hpjQ0UyoEjA5OF-"
        self.page.get(url)

        self.page.wait.doc_loaded()
        time.sleep(3)

        self.page.ele("xpath://div[@class='verifyBtn']").click()
        self.page.wait.doc_loaded()
        time.sleep(4)

        print("为滑块验证码")
        capt_ele = self.page.ele('xpath://*[@id="captcha_modal"]', timeout=2)
        capt_ele.get_screenshot('./element_screenshot.png')

        distance = self.verify(2)
        print(f"滑块距离：{distance}")

        slider_element = self.page.ele(
            "xpath://img[@class='move-img']")

        self.simulate_slider_drag(slider_element, float(distance))
        # 滑块验证处理

        time.sleep(5)
        print(self.page.url, self.page.title)

    def verify(self, type_num):
        """调用云码平台服务"""
        with open('element_screenshot.png', 'rb') as f:
            b = base64.b64encode(f.read()).decode()

        url = "http://api.jfbym.com/api/YmServer/customApi"

        if type_num == 1:
            # 坐标类型
            data = {
                "token": token,
                "type": "30332",
                "direction": "top",
                "click_num": 3,
                "image": b,
            }
        else:
            # 滑块类型
            data = {
                "token": token,
                "type": "22222",
                "image": b,
            }

        _headers = {
            "Content-Type": "application/json"
        }

        response = requests.request("POST", url, headers=_headers, json=data).json()
        print(response)
        return response["data"]["data"]

    def generate_human_track(self, distance):
        """
        生成人类拖动的轨迹
        :param distance: 需要拖动的距离（像素）
        :return: 轨迹点列表，每个点包含(x偏移, y偏移, 延迟时间)
        """
        tracks = []
        current = 0
        mid = distance * 0.7  # 70%处开始减速
        t = 0.2
        v = 0
        move_points = []

        # 第一阶段：加速
        while current < mid:
            a = random.uniform(2, 4)
            v0 = v
            v = v0 + a * t
            move = v0 * t + 0.5 * a * t * t
            current += move
            move_points.append(move)

        # 第二阶段：减速
        while current < distance:
            a = -random.uniform(0.5, 1.5)
            v0 = v
            v = v0 + a * t
            if v < 0.5:  # 防止速度过小
                v = 0.5
            move = v0 * t + 0.5 * a * t * t
            current += move
            move_points.append(move)

        # 添加随机性并生成最终轨迹
        total_points = len(move_points)
        for i, move in enumerate(move_points):
            x_offset = move

            # 添加垂直抖动（模拟手抖）
            if i % random.randint(2, 4) == 0:
                y_offset = random.randint(-2, 2)
            else:
                y_offset = 0

            # 时间间隔（模拟人类反应）
            if i < total_points * 0.3:  # 开始阶段较快
                duration = random.uniform(0.01, 0.03)
            elif i > total_points * 0.7:  # 结束阶段较慢
                duration = random.uniform(0.03, 0.08)
            else:  # 中间阶段
                duration = random.uniform(0.02, 0.05)

            # 随机添加微小停顿
            if random.random() < 0.05:
                duration += random.uniform(0.05, 0.1)

            tracks.append((x_offset, y_offset, duration))

        # 最终微调：到达终点后轻微回拉
        if random.random() < 0.7:
            tracks.append((-random.randint(1, 3), 0, 0.05))

        return tracks

    def simulate_slider_drag(self, slider_element, target_distance):
        """
        模拟人类拖动滑块
        """
        # 移动到滑块并按住
        self.page.actions.move_to(slider_element).hold()
        # 生成轨迹
        tracks = self.generate_human_track(target_distance)
        # 按轨迹拖动
        for track in tracks:
            offset_x, offset_y, duration = track
            self.page.actions.move(offset_x, offset_y, duration=duration / 1000)

        # 释放鼠标
        self.page.actions.release()

    def run(self):
        self.get_page()


if __name__ == '__main__':
    captcga_deal = CaptchaDeal()
    captcga_deal.run()