chore: perp markets token metadata

[shane-moore]

Changes

• For perp markets, we don't have a denom for the base symbol. We just know the base symbol, and we're trying to derive the token metadata from token-metadata package using DenomClient.getDenomToken in the same way that we would if we were trying to derive the metadata via passing the denom. This leads to edge cases and complexity
• In this PR, I propose that simply the flow by grabbing token metadata for derivatives markets "base tokens" via DenomClient.getTokenMetaDataBySymbol and then adding a few required fields to return a proper Token type. Note that it seems like the only useful fields for a derivative base token is the name / symbol / logo properties. The rest don't seem to be used, which is why I feel like the simplification I'm making here with the new getDerivativesToken could be justified

Note

Placed this in draft because we should test some more before merging, but I'm curious your thoughts on this approach

[ThomasRalee]

The implementation looks fine at first glance:

It would be a lot easier to review if we had some test data to benchmark this implementation, i.e the before and after on the list of perp markets we currently support

[bangjelkoski]

The implementation looks fine at first glance:

It would be a lot easier to review if we had some test data to benchmark this implementation, i.e the before and after on the list of perp markets we currently support

I think that this is not the proper approach we should take. We should think where else are we using the symbol based search on the token-metadata and then think of an unified approach that'll take everything into consideration, we should not split the search based on each of the use cases