Rust actix web客户端将负载作为流处理

简单的http客户端，带有rust中的actix。从服务器获取响应，可以处理标头。但当我试图处理有效载荷时，我对生锈的理解失败了。 我的愿望是逐字节处理有效负载，以获得巨大的流和分块数据。简单的body（）解决方案效果很好。但是如何从有效负载中获取字节。它似乎是impl Stream，所以fn size_hint（）是可访问的，但解压器impl Decoder impl Stream不会从流中轮询下一个

梅因

use actix::prelude::*;
use actix_http::encoding::{Decoder};
use actix_web::{HttpMessage,client::ClientBuilder,dev::Payload,http::{Version, StatusCode},web::Bytes};

use futures::future::{TryFutureExt};
use futures_core::stream::Stream;

use std::pin::Pin;
use std::boxed::Box;
use std::task::{Context, Poll};

#[actix_web::main]
async fn main() {
    const URL : [&str; 2] = ["..",".."];

    let client = ClientBuilder::new()
        .disable_timeout()
        .disable_redirects()
        .header("User-Agent", "actix-web/3.0")
        .header("Accept-Encoding", "gzip, deflate, br, chunked")
        .header("Connection", "keepalive")
        .header("DNT", "1")
        .max_http_version( Version::HTTP_2 )
        .finish();

    // Create request builder and send request
    let response = 
    client.get(URL[0]).send().map_err(|err|{
        println!("Error on send {:?}",err);
    });

    match response.await{
        Ok( mut v) => {         
            println!("encoding: {:?}",v.encoding());
            println!("content type: {:?}",v.content_type());
            println!("mime type: {:?}",v.mime_type());
            println!("chunked: {:?}",v.chunked());

            match v.status(){
                StatusCode::OK => {
                    /* works fine
                    match v.body().await {
                        Ok(m) => {  println!("body {:?}", m);   },
                        Err(r) => { println!("Error.body {:?}",r); }
                    }
                    */
                    // but what about handling byte by byte?
                    let payload = v.take_payload();

                    match payload {
                        Payload::None => {  println!(" no data ");  },
                        Payload::H1(h1) => {println!(" data h1 {:?}", h1);  },
                        Payload::H2(d2) => {println!(" data h2 ");          },
                        Payload::Stream(s1) => {
                            println!(" data stream ");
                            
                            println!(" size: {:?}", s1.size_hint());
                            // Decompress impl Decoder
                            // Decoder impl Stream
                            
                            // but how to activate poll_next
                            // compiler can size_hint from Stream
                            // but poll_next not found??
                        }
                    };
                },
                StatusCode::NOT_FOUND => {
                    println!(" resource was not found");
                },
                _ => {
                    
                }
            }
        },
        Err(err) => {
            println!("Error {:?}",err);
        }
    };
}

---

开始时，尝试使用actix:：prelude:：*调出所有必要的定义。这不是解决方案。您会得到什么错误？编译器错误：错误[E0599]：在当前作用域中找不到struct

actix\u http:：encoding:：Decoder

的名为

poll\u next

的方法而不是crate actix web http客户端我使用crate hyper http客户端，它支持分块读取，并使我能够更轻松地控制读取过程。由于没有什么缺点，我必须自己解压数据，使用板条箱异步压缩很容易，默认情况下支持brotli、gzip、bzip2。