Use attestation epoch in beacon fork choice rule

[JustinDrake]

Currently the fork choice rule ignores the attestation epoch. The suggestion (from @djrtwo here) is to replace

if i not in store.latest_messages or target.epoch > store.latest_messages[i].epoch:
    store.latest_messages[i] = LatestMessage(epoch=target.epoch, root=attestation.data.beacon_block_root)

with

attestation_epoch = compute_epoch_of_slot(get_attestation_data_slot(attestation.data))
if i not in store.latest_messages or attestation_epoch > store.latest_messages[i].epoch:
    store.latest_messages[i] = LatestMessage(epoch=attestation_epoch, root=attestation.data.beacon_block_root)

In addition, the suggestion is to add slot to AttestationData.

[mkalinin]

Shouldn't attestation_epoch and attestation.data.target.epoch be the same thing?

According to 0_beacon-chain-validator.md#ffg-vote target.epoch is the current epoch calculated for the state being attested to.

[mkalinin]

It looks like attestation.data.slot is enforced to be of target.epoch since #1509 PR is merged. We may close this issue I guess.

[djrtwo]

yup, good to go