如何使用Actix Web for Rust高效地为大文件服务

我采纳了@Ross Rogers提供的 idea ，并使用了基于Stream的方法.

我使用async_stream crate 来使用stream!()宏来创建流. 然后我声明了一个名为chunk的向量，它基本上告诉我我的块大小应该有多大.

最后，我循环并读取具有声明的块大小的文件，最后生成它们以创建流

以下是我的代码:

#[get("/api/v1/download/file.zip")]
pub async fn download_file_zip(req: HttpRequest, credential: BearerAuth) -> impl Responder {

   let file_path = "path/to/file.zip";

   debug!("File path was set to {}", file_path);

 if let Ok(mut file) = NamedFile::open(file_path.clone()) {

     debug!("File was opened Successfully");

      let my_data_stream = stream! {

        let mut chunk = vec![0u8; 10 * 1024 *1024]; // I decalare the chunk size here as 10 mb 

        loop {

            match file.read(&mut chunk) {

                Ok(n) => {
                    if n == 0 {
                        break;
                    }

                    info!("Read {} bytes from file", n);

                    yield Result::<Bytes, std::io::Error>::Ok(Bytes::from(chunk[..n].to_vec())); // Yielding the chunk here

                }

                Err(e) => {

                    error!("Error reading file: {}", e);
                    yield Result::<Bytes, std::io::Error>::Err(e);
                    break;
                }
            }
        }
    };

    debug!("Sending response...");
    HttpResponse::Ok()
        .content_type("application/octet-stream")
        .streaming(my_data_stream)  // Streaming my response here
} else {
    HttpResponse::NotFound().finish()
  }
}

我暂时接受这一解决方案，直到我找到更好的方法来实现

罗斯·罗杰斯:再次感谢你