Add BuiltinValue type to plutus-core

[ana-pantilie]

Fixes https://github.com/IntersectMBO/plutus-private/issues/1690

[github-actions]

PR Preview Action v1.6.2
🚀 View preview at https://IntersectMBO.github.io/plutus/pr-preview/pr-7225/
Built to branch gh-pages at 2025-09-05 16:28 UTC. Preview will be ready when the GitHub Pages deployment is complete.

[zliu41]

Ready for review. This is a naive implementation, where Value is simply a newtype wrapper around Map ByteString (Map ByteString Integer). The problem is that this doesn't allow efficient access to the size of the largest inner map - it takes O(m) time to do so where m is the size of the outer map. This is likely a problem for costing operations like insertCoin and deleteCoin, which only take O(logm) + O(logk), where k is the size of the largest inner map, since size lookup is more expensive than the operations themselves.

Nonetheless wen can merge this first, and if we need a different representation of Value we can modify it in a separate PR.

[SeungheonOh]

Everything looks reasonable besides some nitpicks.

[SeungheonOh]

we can probably use some better syntax for value. Just using 3tuple for each entry would be more readable.

[zliu41]

That may be slightly better-looking, but then we won't be able to reuse the built-in list parser since a built-in list can't contain a 3-tuple. I'll leave this as future work.

[ana-pantilie]

Can you explain this change? I remember I wasn't sure about the former implementation, so I'd like to understand how this all works.

[zliu41]

knownUni :: uni `Contains` a => uni (Esc a), so it can be instantiated to any uni (Esc a) as long as uni `Contains` a. So there's no need to specify the concrete type.

[effectfully]

I thought we agreed to call it MaryValue? Let's not repeat ledger folks' mistakes

[zliu41]

I don't recall the discussion - do you have a pointer? It's always called Value in plutus, and I'm hesitant to make a name 80% longer unless there's a clear advantage.