feat: Allow individual list of monikers per network

[qwertzlbert]

PR(Pull Request) Overview

A simple change to allow defining individual lists of monikers for each network.

If no monikers variable is defined, the content of the root monikers variable will be used.

This helps in situation were you need to track different monikers across different networks, without the need to define them all globally for all networks

Changes

• Major feature addition or modification
• Bug fix
• Code improvement
• Documentation update

Description of Changes

This change allows to specify individual lists of monikers which should be tracked per configured chain. E.g.

monikers:
  - 'validator1'
  - 'validator2'


  - display_name: 'band'
    chain_id: 'laozi-mainnet'
    monikers:
      - 'validator3'  # <-- will only track 'validator3'
    tracking_addresses:
      - 'xyz'
    nodes:
      - ...


  - display_name: 'injective' # <-- will track 'validator1' and 'validator2'
    chain_id: injective-1
    tracking_addresses:
      - 'xyz'
    nodes:
      - ...

Testing Method

Just a simple execution to test if the overwrite works or not. I did not add any tests yet

[carameleon]

Hi @qwertzlbert
Thank you for your contribution.

Our goal was to simplify the configuration as much as possible. Therefore, currently, if you specify monikers as an array, it will track all chains where those monikers exist.

Your PR seems to be valid for the following scenario:
Let's assume we have:

• chain: A, moniker: abcd
• chain: B, moniker: abcd

And you want to track only abcd from chain A among these two chains, is that correct?

We considered such cases as well, but thought they were rare. Therefore, we chose to specify the monikers in an array and track them if they exist.

[qwertzlbert]

@carameleon Not only that, but in some cases I might want to get data for different validator monikers from different networks to be able to compare them. This is especially useful in scenarios were I track a lot of different chains, but want to track different monikers across them. Currently this would mean I have to define a big list with monikers which also might return data I'm not interested in.
E.g I only want to track my validator and the current leading validator, which might always be a different moniker.

• chain: A, moniker: MyAwesomeValidator, cosmostation
• chain: B, moniker: MyAwesomeValidator, OtherValidator

The change I propose leaves you the option to keep it simple and just define a global list of moniker, the same way as it is now, while still having the flexibility to define an override list for each network if you want.
So it does not really complicate things, but is just a small change providing more flexibility if needed.

Hope this explains the use-case a bit better

[qwertzlbert]

Just found some issues with my change. I'll close the PR until I'm ready

[qwertzlbert]

the global packageMonikers variable needs to be removed in all packages as it created race conditions due to multiple loops writing to this variable at the same time. This worked as long as the moniker variable was the same for all loops, but with my added change which allows moniker configuration per chain, this creates unexpected results.
As the config is passed to the loop via the p variable anyway, p.Monikers is used instead, removing the race-condition

[qwertzlbert]

Found the root cause of the issue faster than I thought so re-opened the PR.