问题描述
是否可以在多个段中发送纯单个http POST请求（不是块编码）？我在考虑使用 httplib.HTTPConnection 并多次调用发送方法（并调用在每个。发送之后，在响应对象上读取
Should it be possible to send a plain, single http POST request (not chunk-encoded), in more than one segment? I was thinking of using httplib.HTTPConnection and calling the send method more than once (and calling read on the response object after each send).
（上下文：I我正在合作设计一个服务器，提供类似于事务的服务[一系列相互关联的请求 - 响应]。我正在寻找最简单，最兼容的HTTP表示。）
(Context: I'm collaborating to the design of a server that offers services analogous to transactions [series of interrelated requests-responses]. I'm looking for the simplest, most compatible HTTP representation.)
推荐答案
在被朋友说服这应该成为可能之后，我找到了一种方法。我重写 httplib.HTTPResponse （nb httplib.HTTPConnection 足够让你指定 response_class 它将实例化。）
After being convinced by friends that this should be possible, I found a way to do it. I override httplib.HTTPResponse (n.b. httplib.HTTPConnection is nice enough to let you specify the response_class it will instantiate).
查看socket.py和httplib.py（特别是 _fileobject.read（） ），我注意到 read（）只允许2件事：
Looking at socket.py and httplib.py (especially _fileobject.read()), I had noticed that read() only allowed 2 things:
- 读取确切的字节数（即使连接未关闭，也会立即返回）
- 读取所有字节，直到连接关闭
我能够扩展这种行为并允许使用几行代码进行免费流式传输。我还必须将HTTPResponse的 will_close 成员设置为0.
I was able to extend this behavior and allow free streaming with just a few lines of code. I also had to set the will_close member of my HTTPResponse to 0.
我仍然有兴趣听到如果这被认为是可接受的或滥用HTTP。
I'd still be interested to hear if this is considered acceptable or abusive usage of HTTP.
这篇关于多次发送httplib.HTTPConnection，多次读取HTTPResponse？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！