-
Notifications
You must be signed in to change notification settings - Fork 185
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Dumping streamed responses breaks later processing #299
Comments
For dumping a response: We have no way to rewind the streamable content so the only way we could responsibly handle that would be to tee it off to an in-memory ByteIO buffer and then replace the underlying file-like object with that so you can handle it as a chunkable response. That said, that will break other assumptions your code has about the response (not taking up all available RAM for example). For dumping a request: We could
|
In case I was not too clear, I am happy with what's happening for requests and I think that the only reasonable course of action is to not dump the body for stream responses. |
👍 I'd be happy to review a PR doing this |
I actually found a way to detect that if the response is being streamed that seems to work in my case. Please check #300. If the approach is ok, I will add a test. |
I am sometimes using
stream=True
and then consuming the raw response chunk by chunk. I if dump one of those responses using toolbelt'sdump()
, it consumes all the content and later processing fails because the download stream is now gone and closed.As for outgoing request data, when a file-like payload is used the body is not dumped, for response bodies this should be achievable too, dumping only the headers.
It appears to me that there is no way to tell that a response is being streamed from the response object itself, so an optional parameters would be needed on dump (I might be wrong).
The text was updated successfully, but these errors were encountered: