feat: support cursed inscriptions

[rafaelcr]

• Keep prev_output and prev_offset in the locations table so we can calculate gaps
• Add /sats/:ordinal/inscriptions endpoint to list all cursed inscriptions for a single sat
• Support new chainhook event cursed_inscription_revealed
• Save and expose curse_type on all endpoints

[github-actions]

Vercel deployment URL: https://ordinals-ghr0hog7v-blockstack.vercel.app 🚀

[lgalabru]

In the deployed version, the inscriptions data are nested in a cursed_inscription_reveal field (vs inscription_reveal).
This is something we could change in future versions, but I think it was a good choice. cursed inscriptions are a bit polarizing, and the ingestion rules are different: you should always wipe cursed inscriptions when ingesting blessed inscription is an idempotent operation.
By keeping the cursed_inscription_reveal schema, developers have to "opt-in" (with an explicit code path) if they want to support cursed inscription.
If we mix everything under inscription_reveal, developers are receiving cursed data by default, and have to "opt-out" by explicitly testing the inscription number if they don't want to support "cursed" inscriptions.

Note that there's also a new field available that you should probably keep: curse_type, which will tell you why the inscription is cursed (even tag, batch, p2wsh as we speak).

[rafaelcr]

@lgalabru a few questions

the inscriptions data are nested in a cursed_inscription_reveal field (vs inscription_reveal)

Does a cursed_inscription_reveal event have the exact same schema as inscription_reveal?

you should always wipe cursed inscriptions when ingesting blessed inscription is an idempotent operation

Can you expand a bit on this? When should I wipe cursed inscriptions?

there's also a new field available that you should probably keep: curse_type

Gotcha, should I return this data on cursed inscription responses too?

[lgalabru]

Does a cursed_inscription_reveal event have the exact same schema as inscription_reveal?

Yep

Can you expand a bit on this? When should I wipe cursed inscriptions?

We can leave this for a future iteration, we're resetting blessed and cursed inscriptions pretty regularly. At some point, we will be adding new curse types, and I think we will keep the blessed inscriptions (to avoid re-ingesting 10M inscriptions), and only reset the cursed inscriptions.

Gotcha, should I return this data on cursed inscription responses too?

That'd be great for the explorer I think.

[lgalabru]

Looking great, thanks @rafaelcr!

[blockstack-devops]

🎉 This PR is included in version 0.3.0 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀