Skip to main content

Module build

Module build 

Source
Expand description

Building proof trees incrementally during trait solving.

This code is a bit of a mess and can hopefully be mostly ignored. For a general overview of how it works, see the comment on ProofTreeBuilder.

Re-exportsยง

use std::marker::PhantomData; ๐Ÿ”’
use derive_where::derive_where; ๐Ÿ”’
use rustc_type_ir as ty; ๐Ÿ”’
use rustc_type_ir::Interner; ๐Ÿ”’
use crate::canonical; ๐Ÿ”’
use crate::delegate::SolverDelegate; ๐Ÿ”’
use crate::solve::Certainty; ๐Ÿ”’
use crate::solve::Goal; ๐Ÿ”’
use crate::solve::GoalSource; ๐Ÿ”’
use crate::solve::QueryResult; ๐Ÿ”’
use crate::solve::inspect; ๐Ÿ”’
use rustc_type_ir::inherent::*; ๐Ÿ”’

Structsยง

EvaluationStepBuilder ๐Ÿ”’
ProofTreeBuilder ๐Ÿ”’
We need to know whether to build a prove tree while evaluating. We pass a ProofTreeBuilder with state: Some(None) into the search graph which then causes the initial EvalCtxt::compute_goal to actually build a proof tree which then gets written into the state.
WipEvaluationStep ๐Ÿ”’
WipProbe ๐Ÿ”’

Enumsยง

WipProbeStep ๐Ÿ”’