Response is not getting compressed or being corrupted? about compress

[ashishnetworks]

Thank you @jonathanong for your comments. I just retested the whole process once again and I realised that my code was already working. I was making mistakes in assuming that response should always be compressed. But what I realised now, after lots of research is that compression works only when the client sends a header with Accept-Encoding. If supported Accept-Encoding is gzip, the response will be compressed in gzip if it's deflate, the response shall be compressed in deflate form, and if it's not mentioned response shall be simple plain data. I was receiving plain text in curl because curl does not send Accept-Encoding in its default request, but when I sent curl request using curl -H 'Accept-Encoding: gzip' -D - http://localhost:3000 my response was coded into compressed form. So my code has been working for always. Thanks

Edit: And you were also right that if content-encoding is set, module will not run
so we don't have to explicitly set the content-encoding, it should be set by compression middleware depending on Accept-Encoding of the request

[jonathanong]

don't set the content encoding urself.

[ashishnetworks]

Thanks, @jonathanong for your time. Well, without setting content encoding myself, I stop getting errors. But I believe it's not getting compressed when I don't set content encoding. Because in such case when I hit the URL using curl I simply get the plain text saying 'Hello World'. If curl doesn't support decompression by default, how am I receiving a plain text then?
And if I don't set encoding myself, the question is, how is it supposed to know which encoding to use out of deflate and gzip?

[jonathanong]

if content-encoding is set, i believe this module will not run.

can you make a PR with a failing test case?

[ashishnetworks]

sorry for being a noob, but what's PR? do you mean sample programs?

[jonathanong]

pull request