Question

我正在尝试使用C＃和ftpwebrequest从ftp服务器下载文件。我可以使用BinaryReader获取字节，但是当我尝试使用br.ReadBytes（int）读取流时，我得到一个BinaryReader不支持搜索操作的错误。

有谁知道如何最好地读取字节，以便我可以将它们写入文件？

这是完整的方法：

    public void DownloadFile(String fileName)
    {
        Logger.Info("starting to download file: " + fileName);

        try
        {
            var downloadFileRequest = (FtpWebRequest)WebRequest.Create(FtpServer + "//" + fileName);
            downloadFileRequest.Credentials = new NetworkCredential(FtpUsername,FtpPassword);
            downloadFileRequest.Method = WebRequestMethods.Ftp.DownloadFile;
            downloadFileRequest.UseBinary = true;

            ServicePoint sp = downloadFileRequest.ServicePoint;
            sp.ConnectionLimit = 2;

            Logger.Info("getting ftp response for download file for " + fileName);

            try
            {
                var downloadResponse = (FtpWebResponse)downloadFileRequest.GetResponse();

                Logger.Info("getting ftp response stream for " + fileName);

                try
                {
                    Stream downloadStream = downloadResponse.GetResponseStream();

                    Logger.Info("File Download status: {0}", downloadResponse.StatusDescription.Trim());

                    Logger.Info("getting binary reader for " + fileName);

                    try
                    {
                       using ( var downloadReader = new BinaryReader(downloadStream))
                       {
                           String destinationFilePath= Path.Combine(LocalFolder, fileName);

                           Logger.Info("writing response stream to " + destinationFilePath);

                           try
                           {
                               using (var downloadWriter = new BinaryWriter(System.IO.File.Open(destinationFilePath, FileMode.Create)))
                               {
                                   downloadWriter.Write(downloadReader.ReadBytes((int)downloadStream.Length));
                               }

                               Logger.Info("successfully saved " + destinationFilePath);

                            }
                            catch (Exception ex)
                            {
                                Logger.Info("could not save " + destinationFilePath+ " b/c: " + ex.Message);
                            }
                        }
                    }
                    catch (Exception ex)
                    {
                        Logger.Info("could not read " + fileName + " b/c: " + ex.Message);
                    }
                }
                catch (Exception ex)
                {
                    Logger.Info("could not open download stream for " + fileName + " b/c: " + ex.Message);
                }
                finally
                {
                    downloadResponse.Close();
                }
            }
            catch (Exception ex)
            {
                Logger.Info("could not get ftp response stream for " + fileName + " b/c: " + ex.Message);
            }
        }
        catch (Exception ex)
        {
            Logger.Info("could not get ftp request stream for " + fileName + " b/c: " + ex.Message);
        }
    }

这是正在进行的服务的一部分，所以我不想抛出会阻止服务的错误。相反，我正在写日志。以下是此方法的日志内容：

2009-10-07 16:33:29.1421|INFO|xxx.Web.Controllers.FtpController|starting to download file: 2009-10-06155728Z_metadata.txt
2009-10-07 16:33:29.1421|INFO|xxx.Web.Controllers.FtpController|getting ftp response for download file for 2009-10-06155728Z_metadata.txt
2009-10-07 16:33:29.6661|INFO|xxx.Web.Controllers.FtpController|getting ftp response stream for 2009-10-06155728Z_metadata.txt
2009-10-07 16:33:29.6661|INFO|xxx.Web.Controllers.FtpController|File Download status: 125 Data connection already open; Transfer starting.
2009-10-07 16:33:29.6721|INFO|xxx.Web.Controllers.FtpController|getting binary reader for 2009-10-06155728Z_metadata.txt
2009-10-07 16:33:29.6721|INFO|xxx.Web.Controllers.FtpController|writing response stream to C:\\Resumes\\2009-10-06155728Z_metadata.txt
2009-10-07 16:33:29.6951|INFO|xxx.Web.Controllers.FtpController|could not save C:\\Resumes\\2009-10-06155728Z_metadata.txt b/c: This stream does not support seek operations.

我一直在这方面工作太久了，所以任何帮助都会受到赞赏！

谢谢！

Answer 1

你不应该依赖Stream.Length，这可能是错误的。您只需要读取while循环中的所有字节，直到没有更多的字节要读取。

MemoryStream ms = new MemoryStream();
byte[] chunk = new byte[4096];
int bytesRead;
while ((bytesRead = downloadStream.Read(chunk, 0, chunk.Length)) > 0)
{
     ms.Write(chunk, 0, bytesRead);
}

从那里开始，所有读取的数据都在MemoryStream中，您可以使用它初始化BinaryReader。

Answer 2

Darkassassin的答案很棒！这是我在看到他的帖子之前最终开始工作的代码。它的b / c略有不同，它直接写入文件而不是内存流。

我替换了这一行：

downloadWriter.Write(downloadReader.ReadBytes((int)downloadStream.Length));

用这个：

var buffer = new byte[BufferSize];
int readCount = downloadStream.Read(buffer, 0, BufferSize);
while (readCount > 0)
{
    downloadWriter.Write(buffer, 0, readCount);
    readCount = downloadStream.Read(buffer, 0, BufferSize);
}

（BufferSize = 4096）

再次感谢您的快速帮助！

C＃BinaryReader“流不支持搜索操作”

2 个答案: