Python爬取动态加载内容的隐藏电话号码：API请求方法详解

import re

def extract_post_id_from_url(url):
    """
    从Haraj网站的URL中提取postId。
    示例: "https://haraj.com.sa/1194697687" -> 94697687
    """
    match = re.search(r'/(\d+)$', url)
    if match:
        full_id_str = match.group(1)
        # 根据观察，postId是URL末尾数字去除前两位"11"后的部分
        if full_id_str.startswith('11') and len(full_id_str) > 2:
            return int(full_id_str[2:])
        return int(full_id_str) # 如果没有"11"前缀，则直接返回
    return None

# 示例用法
target_url = "https://haraj.com.sa/1194697687"
post_id = extract_post_id_from_url(target_url)
if post_id:
    print(f"提取到的 Post ID: {post_id}")
else:
    print("无法从URL中提取 Post ID。")

import requests
import re

def extract_post_id_from_url(url):
    """
    从Haraj网站的URL中提取postId。
    示例: "https://haraj.com.sa/1194697687" -> 94697687
    """
    match = re.search(r'/(\d+)$', url)
    if match:
        full_id_str = match.group(1)
        # 根据观察，postId是URL末尾数字去除前两位"11"后的部分
        if full_id_str.startswith('11') and len(full_id_str) > 2:
            return int(full_id_str[2:])
        return int(full_id_str) # 如果没有"11"前缀，则直接返回
    return None

def get_hidden_phone_number(page_url):
    """
    通过模拟API请求获取隐藏的电话号码。
    """
    # 1. 从页面URL中提取 postId
    post_id = extract_post_id_from_url(page_url)
    if post_id is None:
        print(f"错误: 无法从URL '{page_url}' 中提取有效的 postId。")
        return None

    # 2. 定义API请求参数
    api_url = "https://graphql.haraj.com.sa"

    # 根据浏览器网络分析，这些参数可能为空或特定值
    params = {
        "queryName": "postContact",
        "token": "",
        "clientId": "",
        "version": ""
    }

    headers = {
        "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.0.0 Safari/537.36",
        "accept": "*/*",
        "accept-language": "en-US,en;q=0.9",
        "content-type": "application/json",
        "origin": "https://haraj.com.sa",
        "referer": page_url # 设置referer以模拟真实浏览器行为
    }

    # GraphQL请求的Payload
    payload = {   
        "query": "query postContact($postId: Int!) {postContact(postId: $postId){contactText}}",
        "variables": {
            "postId": post_id # 使用动态获取的postId
        }
    }

    try:
        # 3. 发送POST请求
        response = requests.post(api_url, params=params, headers=headers, json=payload)
        response.raise_for_status() # 检查HTTP请求是否成功

        # 4. 解析JSON响应
        data = response.json()

        # 提取电话号码
        contact_text = data.get('data', {}).get('postContact', {}).get('contactText')
        return contact_text

    except requests.exceptions.RequestException as e:
        print(f"请求失败: {e}")
        return None
    except KeyError as e:
        print(f"解析响应数据失败: 缺少键 {e}")
        return None

# 主函数调用示例
if __name__ == "__main__":
    target_page_url = "https://haraj.com.sa/1194697687" # 替换为实际的页面URL
    phone_number = get_hidden_phone_number(target_page_url)

    if phone_number:
        print(f"成功获取到电话号码: {phone_number}")
    else:
        print("未能获取电话号码。")

    # 另一个示例URL
    another_page_url = "https://haraj.com.sa/1199808969"
    phone_number_2 = get_hidden_phone_number(another_page_url)
    if phone_number_2:
        print(f"成功获取到另一个电话号码: {phone_number_2}")
    else:
        print("未能获取另一个电话号码。")