Safe Haskell | Safe-Inferred |
---|---|
Language | Haskell2010 |
Monadic interface to e-graph stateful computations
Synopsis
- egraph :: Language l => EGraphM l a -> (a, EGraph l)
- represent :: Language l => Fix l -> EGraphM l ClassId
- add :: Language l => ENode l -> EGraphM l ClassId
- merge :: Language l => ClassId -> ClassId -> EGraphM l ClassId
- rebuild :: Language l => EGraphM l ()
- canonicalize :: Functor l => ENode l -> EGraph l -> ENode l
- find :: ClassId -> EGraph l -> ClassId
- emptyEGraph :: Language l => EGraph l
- type EGraphM s = State (EGraph s)
- runEGraphM :: EGraph l -> EGraphM l a -> (a, EGraph l)
- data EGraph l
- modify :: forall (m :: Type -> Type) s. Monad m => (s -> s) -> StateT s m ()
- get :: forall (m :: Type -> Type) s. Monad m => StateT s m s
- gets :: forall (m :: Type -> Type) s a. Monad m => (s -> a) -> StateT s m a
Documentation
egraph :: Language l => EGraphM l a -> (a, EGraph l) Source #
Run EGraph computation on an empty e-graph
Example
egraph $ do id1 <- represent t1 id2 <- represent t2 merge id1 id2
represent :: Language l => Fix l -> EGraphM l ClassId Source #
Represent an expression (Fix l
) in an e-graph by recursively
representing sub expressions
merge :: Language l => ClassId -> ClassId -> EGraphM l ClassId Source #
Merge two e-classes by id
E-graph invariants may be broken by merging, and rebuild
should be used
eventually to restore them
rebuild :: Language l => EGraphM l () Source #
Rebuild: Restore e-graph invariants
E-graph invariants are traditionally maintained after every merge, but we
allow operations to temporarilly break the invariants (specifically, until we call
rebuild
)
The paper describing rebuilding in detail is https://arxiv.org/abs/2004.03082
canonicalize :: Functor l => ENode l -> EGraph l -> ENode l Source #
Canonicalize an e-node
Two e-nodes are equal when their canonical form is equal. Canonicalization makes the list of e-class ids the e-node holds a list of canonical ids. Meaning two seemingly different e-nodes might be equal when we figure out that their e-class ids are represented by the same e-class canonical ids
canonicalize(𝑓(𝑎,𝑏,𝑐,...)) = 𝑓((find 𝑎), (find 𝑏), (find 𝑐),...)
find :: ClassId -> EGraph l -> ClassId Source #
Find the canonical representation of an e-class id in the e-graph Invariant: The e-class id always exists.
emptyEGraph :: Language l => EGraph l Source #
The empty e-graph. Nothing is represented in it yet.
E-graph stateful computations
runEGraphM :: EGraph l -> EGraphM l a -> (a, EGraph l) Source #
Run EGraphM
computation on a given e-graph
E-graph definition re-export
E-graph representing terms of language l
.
Intuitively, an e-graph is a set of equivalence classes (e-classes). Each e-class is a set of e-nodes representing equivalent terms from a given language, and an e-node is a function symbol paired with a list of children e-classes.