feat: Adds ws and wss multiaddr

[gpestana]

Adds ws and wss multiaddr as defined in https://github.com/multiformats/multiaddr/blob/master/protocols.csv

Related to #64

[Stebalien]

The WS protocol is currently defined in go-ws-transport (but we don't define the wss protocol because we can't use it).

However, I understand the issue.

@whyrusleeping

Issue: We can neither parse nor print a multiaddr unless we understand the full multiaddr.
Solutions:

1. This PR (put every protocol definition we can in one repo).
2. Add an "unknown" protocol to the string format: Handle unparsable multiaddrs multiaddr#70

The second completely solves the issue that JS is having but the first one is (a) easier and (b) provides better usability.

[gpestana]

The 2) option sounds reasonable to me, although I might not fully understand what are the issues JS is having with the option 1). I think that if the WS protocol is implemented somewhere else, it should be documented somewhere here as it gets confusing for newcomers.

What would be the criteria to add a new protocol to the list and implement it vs treat it as unknown?

[Stebalien]

The current issue for javascript land is that go-multiaddr doesn't understand protocols like wss or webrtc (?). This means that:

1. We throw away these addresses when we encounter them (in the DHT).
2. We can't dial addresses like /ip4/4.3.2.1/tcp/321/p2p/QmRelay/p2p-circuit/ip4/1.2.3.4/tcp/123/wss/ipfs/QmDest even if we speak all the protocols we need to speak (those before p2p-circuit).

Adding these protocols to this package will fix these issues for now. However, the second we add a new protocol, we'll have the same problem.

We can fix 1 by simply not decoding these addresses. However, that would be a large, weird change and I'd like to avoid it if possible. Note: that issue will be fixed eventually once we get ipfs/notes#291 (comment) but that's a ways off.

Fix 2 (the "unknown" protocol) fixes issue 2 from above. That's basically the only fix I can think of.

[Stebalien]

We went with option 1 (define everything here) for now. Updated PR in #122.