Loading...
C
h
alk-RLS integration
active work items
remove concept of
“local
and external” and reformulate in terms of crate-ids — make coherence relative to a specific crate?
@
Sunjay V
refactor lowering code to salsa queries that
“demand”
information
@
Nicholas M
@
Alexandre M
integration with RLS / test runner can use
this callback setup
.
how to handle ids?
@
Wilco K
(detrumi)
chalk uses ids, RLS has a distinct notion of id
per definition
list out the various places that chalk uses ids
compare to corresponding RLS types
refactor to match
error types
should we add an error type? perhaps it unifies to
CannotProve
or something like that?
discussion here
refactor RLS2.0 Ty
@
Florian D
morph it into type, isomorphic to chalk::Ty bit by bit, than swap out the impl
things in the RLS
Some links into the RLS HIR:
data from an impl
TypeRef
type
defines types
target_ty
method
shows how to combine name resolution
named resolved type
Things to add:
TraitRef
type for resolved traits
representation of bounds
refactor type representation to be closer to chalk’s
consolidate variants into
Apply
remove stuff like names and implement proper pretty-printing
overall flow
main integration point is probably a
prove
query that, given a canonical goal, gives back the result
discussion here
chalk-engine will call back to RLS to request
“stuff
from impls”
this will invoke chalk’s lowering code
we might use this to avoid doing eager name resolution for everything
Please turn on JavaScript to use Paper in all of its awesomeness. ^_^
active work items
things in the RLS
overall flow