2.1 KiB
Compilation and readback
How are terms compiled to interaction net nodes?
HVM-Core has a bunch of useful nodes to write IC programs.
Every node contains one main
port 0
and two auxiliary
ports, 1
and 2
.
There are 6 kinds of nodes, Eraser, Constructor, Reference, Number, Operation and Match.
A lambda λx x
compiles into a Constructor node.
An application ((λx x) (λx x))
also compiles into a Constructor node.
0 - Points to the lambda occurrence 0 - Points to the function
| |
λ Lambda @ Application
/ \ / \
1 2 - Points to the lambda body 1 2 - Points to the application occurrence
| |
Points to the lambda variable Points to the argument
When reading back, if we visit a Constructor via port 0 then we know it's a lambda, and if we visit it via port 2 it's an application.
- The
Number
node uses the label to store it's number. - An
Op2
node uses the label to store it's operation. - And a
Constructor
node can have a label too! This is used fordup
and [lambda tags](automatic-vectorization-with-tagged-lambdas.md.
A duplication dup a b = x
compiles into a Constructor node too, but with a different label.
A superposition {a b}
compiles to a Constructor node too. The difference here comes from context too.
Additionally, nodes have labels. We use the label to store data in the node's memory, which can be used for various purposes.
0 - Points to the sup occurence 0 - Points to the duplicated value
| |
# Superposition # Duplication
/ \ / \
1 2 - Points to the second value 1 2 - Points to the second binding
| |
Points to the first value Points to the first binding
Check out HVM-Core, one of the Higher Order Company's projects, to know more about this.