Add get level from hexagon

[melrief]

No description provided.

[ghost]

For this you could generate random characters from '0' to '9' and then use the String-argument BigInt constructor.

[melrief]

I'm not sure but probably summing BigInts is faster than parsing String

[melrief]

@shishkin can you have a look at this? I would like to use this code.

[shishkin]

Thanks for your effort here. I'm not sure exposing level on the Encoding is a good idea. I tried to make Encoding as dumb and single-purpose as possible - it just encodes/decodes Zone objects to/from an arbitrary representation. Zone already knows about its level: TeraHex.decode(code).level should already do that. Isn't this enough or usable in your case?

[shishkin]

At this point, API of the library is in complete flux and every change is potentially breaking. Especially around JS/JVM interop. I'd refrain from giving a false sense of compatibility and stick to 0.1.x for now.

[melrief]

The reason why I and @ilyabo in particular are not using TeraHex.decode is because it's slow. For interactive applications it's not fine. Having a faster way to get the level of an hexagon can improve the performance of most of the applications. The dashboard, for instance, uses the length of the hexcode minus four. This PR is just moving this functionality from Ilya's code to the library so that also the backend can use it.

[shishkin]

Then I'd rather make Zone decoding faster. Is it slow on JVM too? Can you elaborate on your scenario a bit? Are you getting levels of multiple hexagons in a tight loop? Are they expected to have different levels? Then I can try to micro-benchmark that scenario and see what is the bottleneck. Also, what would acceptable performance be?

[ilyabo]

Honestly, I don't think this tiny improvement is worth spending much time in discussions :) Anyways, here's a quick benchmark:

> var t=require('terahex'); console.time('decode');  for (var i=0; i< 1000000; i++) t.decode('1300123123123'+(i%9)).level; console.timeEnd('decode')
decode: 8762ms

> var t=require('terahex'); function level(code) {return code.length - 4}; console.time('strlen');  for (var i=0; i< 1000000; i++) level('130012312312'+(i%9)); console.timeEnd('strlen')
strlen: 13ms

I also think Encoding is the right place for it, because this knowledge is part of the encoding scheme.

If we don't have this function in terahex, we'll end up using the helper functions on both Scala and JS sides for that, as now, which is a bad thing, because it's part of the implementation detail which we shouldn't rely upon and even worse have it reimplemented in multiple places.