Bulk token metadata endpoint

[heidar-stacks]

Fetching token (FT and NFT) metadata currently requires one request per token ID, which causes overhead when displaying lists (e.g. in address page > holdings section and tokens table). This results in excessive network calls and potential rate limit.

Proposed solution: adding a new bulk metadata endpoint:

GET /metadata/v1/ft?ids=ID1,ID2,ID3...

potential issue with hitting url max size

POST /metadata/v1/ft
Body: { ids: [ "ID1", "ID2", "ID3" ] }

better in handling larger lists but potential issue with caching

related to: stx-labs/explorer#2544

[rafaelcr]

gm @heidar-stacks , we've looked over the issue on our end and as discussed in our call here's a strategy that would work for what you want to achieve:

1. Create a new endpoint /metadata/search?contract=&contract=&contract=... that accepts a (required) array param called contract via the query string. This param would accept contract ID values and optional token numbers, for example:
   • SM3VDXK3WZZSA84XXFKAFAF15NNZX32CTSG82JFQ4.sbtc-token
   • SP3D6PV2ACBPEKYJTCMH7HEN02KP87QSP8KTEH335.nakapack-nft
   • SP3D6PV2ACBPEKYJTCMH7HEN02KP87QSP8KTEH335.nakapack-nft:120
2. A query will be then built that checks the metadata table looking for entries based on those contract principals and optional token numbers. If a token number is not specified, it should default to 1.
3. The endpoint will then return all metadata entries it has found for the given contracts, but should take into account when some contract cannot be found or when none of them can be found. If you want you could return a compact metadata result instead of the whole thing e.g. only the images, name, symbol, etc. instead of the entire SIP-016 parsed entry.

The only limiting factor for this endpoint is the max size of the URI that is generally allowed in the web which is 32KB, but this should be OK if you divide your calls by 10-20 endpoints per call or something like that.

What do you think? I would love if you could implement this like you suggested because we're really short on devs at the moment. Please let me know if you have any questions!