如何通过chunck字节数组将chunck转换为base64字符串？

Question

我们正在阅读文件的字节数组，并将其转换成字符串的base64如下

public static string ZipToBase64() 
      { 
       FileUpload fileCONTENT = FindControl("FileUploadControl") as FileUpload; 

       byte[] byteArr = fileCONTENT.FileBytes; 

       return Convert.ToBase64String(byteArr); 

      } 

    string attachmentBytes = ZipToBase64(); 
    string json1 = "{ \"fileName\": \"Ch01.pdf\", \"data\": " + "\"" + attachmentBytes + "\"}"; 

当我们尝试将大型文件转换高达1 GB为base64，字符串，则抛出内存异常。我们将这个json发送给restful wcf服务。以下是我在RESTful WCF服务中的方法。

[OperationContract] 
     [WebInvoke(Method = "POST", BodyStyle = WebMessageBodyStyle.WrappedRequest, RequestFormat = WebMessageFormat.Json, ResponseFormat = WebMessageFormat.Json)] 
     public void UploadFile1(Stream input) 
     { 

      string UserName = HttpContext.Current.Request.Headers["UserName"]; 
      string Password = Sql.ToString(HttpContext.Current.Request.Headers["Password"]); 
      string sDevideID = Sql.ToString(HttpContext.Current.Request.Headers["DeviceID"]); 
      string Version = string.Empty; 
      if (validateUser(UserName, Password, Version, sDevideID) == Guid.Empty) 
      { 
       SplendidError.SystemWarning(new StackTrace(true).GetFrame(0), "Invalid username or password for " + UserName); 
       throw (new Exception("Invalid username or password for " + UserName)); 
      } 


      string sRequest = String.Empty; 
      using (StreamReader stmRequest = new StreamReader(input, System.Text.Encoding.UTF8)) 
      { 
       sRequest = stmRequest.ReadToEnd(); 
      } 
      // http://weblogs.asp.net/hajan/archive/2010/07/23/javascriptserializer-dictionary-to-json-serialization-and-deserialization.aspx 

      JavaScriptSerializer json = new JavaScriptSerializer(); 
      // 12/12/2014 Paul. No reason to limit the Json result. 
      json.MaxJsonLength = int.MaxValue; 

      Dictionary<string, string> dict = json.Deserialize<Dictionary<string, string>>(sRequest); 
      string base64String = dict["data"]; 
      string fileName = dict["fileName"]; 


      byte[] fileBytes = Convert.FromBase64String(base64String); 
      Stream stream = new MemoryStream(fileBytes); 

      //FileStream fs1 = stream as FileStream; 

      string networkPath = WebConfigurationManager.AppSettings["NetWorkPath"]; 
      File.WriteAllBytes(networkPath + "/" + fileName, fileBytes); // Requires System.IO 
     } 

请转换大字节数组转换为字符串的base64

Answer 1

使用流输入的事实在WCF服务并不真正意味着你传递流的方式提供任何解决方案。事实上在你的情况下，你不这样做，因为：

1. 你在内存中的客户端缓冲整个文件来建立json字符串。
2. 您通过stmRequest.ReadToEnd()将整个文件缓冲在内存中的服务器上。

所以不会发生流式传输。首先你应该认识到这里不需要json - 你只需要在你的http请求体中传递文件即可。你应该做的首先是扔掉下面的安全检查所有代码在你UploadFile1方法，而是这样做：

public void UploadFile1(string fileName, Stream input) { 
    // check your security headers here 
    string networkPath = WebConfigurationManager.AppSettings["NetWorkPath"]; 
    using (var fs = File.Create(networkPath + "/" + fileName)) { 
     input.CopyTo(fs); 
    } 
} 

在这里，我们只是没有任何缓冲复制输入流输出流（文件）（当然CopyTo将缓冲大块，但它们会很小）。马克您的服务方法：

[WebInvoke(Method = "POST", UriTemplate = "/UploadFile1/{fileName}")] 

为了使您在查询字符串传递的文件名。

现在到客户端。不确定您使用哪种方法与服务器进行通信，我将使用原始HttpWebRequest显示示例。

var filePath = "path to your zip file here"; 
var file = new FileInfo(filePath); 
// pass file name in query string 
var request = (HttpWebRequest)WebRequest.Create("http://YourServiceUrl/UploadFile1/" + file.Name);    
request.Method = "POST"; 
// set content length 
request.ContentLength = file.Length; 
// stream file to server 
using (var fs = File.OpenRead(file.FullName)) { 
    using (var body = request.GetRequestStream()) { 
     fs.CopyTo(body); 
    } 
} 
// ensure no errors 
request.GetResponse().Dispose();