feat: add more recursion endpoints

[janniks]

Adding these endpoints covered by ord
https://docs.ordinals.com/inscriptions.html?highlight=rec#recursion

A couple other endpoints that inscriptions may access are the following:

/blockheight: latest block height.
/blockhash: latest block hash.
/blockhash/<HEIGHT>: block hash at given block height.
/blocktime: UNIX time stamp of latest block.

---

todo

• cacheing: what's your preferred etag/cacheing option? we could add a stats/inscriptioncountperblock cache for most, and a custom new cache for the /:blockheight show endpoint.

[github-actions]

Vercel deployment URL: https://ordinals-5ut3rkhrx-blockstack.vercel.app 🚀

[codecov]

Codecov Report

Merging #144 (a3c3add) into develop (a7674a9) will decrease coverage by 4.25%.
The diff coverage is 100.00%.

❗ Current head a3c3add differs from pull request most recent head 389ec23. Consider uploading reports for the commit 389ec23 to get more accurate results

@@             Coverage Diff             @@
##           develop     #144      +/-   ##
===========================================
- Coverage    91.98%   87.74%   -4.25%     
===========================================
  Files           26       34       +8     
  Lines         2970     3622     +652     
  Branches       239      294      +55     
===========================================
+ Hits          2732     3178     +446     
- Misses         236      440     +204     
- Partials         2        4       +2     

Impacted Files	Coverage Δ
src/api/init.ts	70.49% <100.00%> (+1.52%)	⬆️
src/api/routes/recursion.ts	100.00% <100.00%> (ø)
src/api/schemas.ts	100.00% <100.00%> (ø)
src/pg/pg-store.ts	99.27% <100.00%> (+1.79%)	⬆️

... and 19 files with indirect coverage changes

[rafaelcr]

Looking good! A few comments

[rafaelcr]

Is the word blockheight the right thing to return if we don't have a proper block height here? Not familiar with the ord docs on this

[janniks]

Yep, that's to match the current ord implementation. https://sourcegraph.com/search?q=context:global+repo:ordinals/ord+and+%7C%7C+%22

[rafaelcr]

I think we could make these queries more efficient if we modified the chain_tip materialized view to keep also the block hash and timestamp. That way it would be a O(1) retrieval from the DB always.

[janniks]

Forgot about this one. A lot of recursive viewing doesn't really work in our explorer due to security concerns, so it might not be too important to have these endpoints right now..