Self-Vote for Staking (among others)

[kianenigma]

Treats some leftover review from #1915.

• Adds an automatic self-vote to each validator. This is an entirely artificial edge in Phragmen and will be removed at the end of the algorithm (meaning that it is NOT counted when slashing/rewarding).
• Improves the performance by removing a layer of iteration in every inner loop of the main algorithm (--at the expense of storing two sets of indexes in each Edge). This basically makes the complexity of this implementation exactly the same when compared to the reference implementation.
• Adds the Phragmen equalise() post-processing, the parameters of which are currently configured statically in the code and NOT via a storage item (I think it will be constant --should refactor to a storage item if not).
• Some other minor fixes.
• Few more tests to verify the above.

Things that I am not satisfied with

• How to react to potential overflows.
• Maybe storing the direct reference of a Candidate in Edge is better, given that it will probably create a mess of lifetime.

[parity-cla-bot]

It looks like @kianenigma signed our Contributor License Agreement. 👍

Many thanks,

Parity Technologies CLA Bot

[parity-cla-bot]

It looks like @kianenigma signed our Contributor License Agreement. 👍

Many thanks,

Parity Technologies CLA Bot