Add permessage-deflate support, again

[SvizelPritula]

In #328, @kazk implemented support for the permessage-deflate extention. Implementing this extention required the ability to parse the Sec-Websocket-Extentions header, so @kazk submitted a pull request to the headers crate. However, the maintainers of the headers crate don't wan't to support this header yet, as they don't wan't to commit to any particular API. As such, they suggested that it should be implemented in this or some other crate, tweaked as necessary and potentially moved into the headers crate in the future.

This PR reverts the commit that reverted @kazk's commit that added permessage-deflate support. It also copies his implementation of SecWebsocketExtentions intended for the headers crate.

The implementation of SecWebsocketExtentions relied on some internal utilities of the headers crate. I've removed some of those dependencies and re-implemented others. I have also gated the implementation behind the existing handshake feature, as well as any code that relies on it, since it relies on the headers crate, which in turn relies on many other HTTP crates.

Blocking issues

• @nakedible-p found an error in the implementation. As I know very little about the Websocket spec, I'm unsure if I can fix it properly.
• The headers crate is licensed under the MIT license, while this crate uses a dual MIT+Apache license. Assuming we don't want to change the license on this crate, @kazk would need to consent to the re-licensing of his PR.

Unresolved questions

• It is impossible to use the deflate feature without having tungstenite handle the Websocket handshake. There is a WebSocket::from_raw_socket_with_extensions method to allow for exactly that, but it takes an Extensions struct, which has no public constructor.
• The only way to get such a struct is from the WebSocketConfig::accept_offers method, which implements the server part of extension negotiation and requires the handshake feature. This method is intended to allow for tungstenite to be integrated into web frameworks. No similar public method exists for clients, although that might not be a big issue, since a client knows whether a request will be a Websocket request up front.
• To me, the WebSocketConfig::accept_offers feels out of place on the WebSocketConfig. We could make it a standalone function, perhaps in the extensions module.
• I have made the SecWebsocketExtentions implementation public. This is necessary to make the WebSocketConfig::accept_offers method usable. This does mean that if SecWebsocketExtentions were to be added to headers in the future, switching to that implementation would (I think) be a breaking change. An alternative would be to have WebSocketConfig::accept_offers take and return a HeaderValue, which it would parse itself. This would allow us to make SecWebsocketExtentions a private implementation detail.
• This is a breaking change, since a new field was added to WebSocketConfig and a new variant was added to the Error enum. Since this field and variant depends on the deflate feature, I've annotated both with #[non_exhaustive]. I've done the same to ProtocolError, which currently has variants gated behind handshake. This sadly means that you cannot create WebSocketConfig using the initializer syntax, instead you have to create an instance with WebSocketConfig::default and mutate it afterward. Sadly, I see no way to avoid this. It would be possible to disable #[non_exhaustive] if all relevant features are enabled.

[nakedible-p]

I will be happy to go through the details of the permessage-deflate spec – or rather even fix the bug myself, while some missing window bits support. However, the bug is extremely niche as these extension fields are very rarely used, so I'd prefer to try to get this merged and then do an improvement pull on top of that.