如何避免 HttpClient 丢失请求头：通过 HttpRequestMessage 解决并优化

using System.Net.Http;
using System.Text;
using Newtonsoft.Json;
using Newtonsoft.Json.Serialization;

namespace ConsoleApp9
{
    internal class Program
    {
        private static readonly JsonSerializerSettings serializerSettings = new JsonSerializerSettings
        {
            ContractResolver = new CamelCasePropertyNamesContractResolver(),
            NullValueHandling = NullValueHandling.Ignore
        };

        private static readonly HttpClient httpClient = new HttpClient(); // 复用HttpClient实例
        private static readonly SemaphoreSlim semaphore = new SemaphoreSlim(100); // 限制并发请求数量为100

        static async Task Main(string[] args)
        {
            List<Task> tasks = new List<Task>();
            int taskNoCounter = 1; // 用于跟踪 taskno
            // 只使用一个HttpClient对象（全局共享）
            for (int i = 0; i < 50; i++)
            {
                tasks.Add(Task.Run(async () =>
                {
                    // 等待信号量，控制最大并发数
                    await semaphore.WaitAsync();

                    try
                    {
                        var postData = new
                        {
                            taskno = taskNoCounter++,
                            content = "等待翻译的内容"
                        };
                        var json = JsonConvert.SerializeObject(postData, serializerSettings);
                        var reqdata = new StringContent(json, Encoding.UTF8, "application/json");

                        // 设置请求头语言
                        httpClient.DefaultRequestHeaders.Add("Accept-Language", "en-US");                      
                        // 发送请求
                        var result = await httpClient.PostAsync("http://localhost:5000/translate", reqdata);

                        // 读取并反序列化 JSON 数据
                        var content = await result.Content.ReadAsStringAsync();
                        var jsonResponse = JsonConvert.DeserializeObject<Response>(content);
                        var response = jsonResponse.Data.Content;
                       
                        // 反序列化后，直接输出解码后的文本
                        Console.WriteLine($"结果为：{response}");
                    }
                    catch (Exception ex)
                    {
                        Console.WriteLine($"请求失败: {ex.Message}");
                    }
                    finally
                    {
                        // 释放信号量
                        semaphore.Release();
                    }
                }));
            }

            await Task.WhenAll(tasks);
        }
    }

    // 定义与响应结构匹配的类
    public class Response
    {
        public int Code { get; set; }
        public ResponseData Data { get; set; }
        public string Msg { get; set; }
    }

    public class ResponseData
    {
        public string Content { get; set; }
        public string Lang { get; set; }
        public int Taskno { get; set; }
    }
}

from flask import Flask, request, jsonify
from google.cloud import translate_v2 as translate

app = Flask(__name__)

# 初始化 Google Cloud Translate 客户端
translator = translate.Client()

@app.route('/translate', methods=['POST'])
def translate_text():
    try:
        # 从请求中获取 JSON 数据
        data = request.get_json()

        # 获取请求的文本内容
        text = data.get('content')
        taskno = data.get('taskno', 1)

        # 获取请求头中的 Accept-Language 信息，默认为 'zh-CN'
        accept_language = request.headers.get('Accept-Language', 'zh-CN')

        # 调用 Google Translate API 进行翻译
        result = translator.translate(text, target_language=accept_language)

        # 构造响应数据
        response_data = {
            "code": 200,
            "msg": "OK",
            "data": {
                "taskno": taskno,
                "content": result['translatedText'],
                "lang": accept_language
            }
        }

        # 返回 JSON 响应
        return jsonify(response_data), 200

    except Exception as e:
        return jsonify({"code": 500, "msg": str(e)}), 500


if __name__ == "__main__":
    app.run(debug=True, host="0.0.0.0", port=5000)