refactor(core): Standardized task graph

datafusion-optd-cli/src/main.rs

@@ -156,7 +156,7 @@ pub async fn main() -> Result<()> {
     let args = Args::parse();
 
     tracing_subscriber::fmt()
-        .with_max_level(tracing::Level::INFO)
+        .with_max_level(tracing::Level::TRACE)
         .with_target(false)
         .with_ansi(false)
         .init();

optd-core/src/cascades/optimizer.rs

@@ -3,24 +3,25 @@
 // Use of this source code is governed by an MIT-style license that can be found in the LICENSE file or at
 // https://opensource.org/licenses/MIT.
 
-use std::collections::{BTreeSet, HashMap, HashSet, VecDeque};
+use std::collections::{BTreeSet, HashMap, HashSet};
 use std::fmt::Display;
+use std::sync::atomic::{AtomicUsize, Ordering};
 use std::sync::Arc;
 
 use anyhow::Result;
 use tracing::trace;
 
 use super::memo::{ArcMemoPlanNode, GroupInfo, Memo};
-use super::tasks::OptimizeGroupTask;
 use super::{NaiveMemo, Task};
 use crate::cascades::memo::Winner;
+use crate::cascades::tasks::get_initial_task;
 use crate::cost::CostModel;
 use crate::nodes::{
     ArcPlanNode, ArcPredNode, NodeType, PlanNodeMeta, PlanNodeMetaMap, PlanNodeOrGroup,
 };
 use crate::optimizer::Optimizer;
 use crate::property::{PropertyBuilder, PropertyBuilderAny};
-use crate::rules::Rule;
+use crate::rules::{Rule, RuleMatcher};
 
 pub type RuleId = usize;
 
@@ -43,11 +44,19 @@ pub struct OptimizerProperties {
 
 pub struct CascadesOptimizer<T: NodeType, M: Memo<T> = NaiveMemo<T>> {
     memo: M,
-    pub(super) tasks: VecDeque<Box<dyn Task<T, M>>>,
+    /// Stack of tasks that are waiting to be executed
+    tasks: Vec<Box<dyn Task<T, M>>>,
+    /// Monotonically increasing counter for task invocations
+    task_counter: AtomicUsize,
     explored_group: HashSet<GroupId>,
     explored_expr: HashSet<ExprId>,
-    fired_rules: HashMap<ExprId, HashSet<RuleId>>,
-    rules: Arc<[Arc<dyn Rule<T, Self>>]>,
+    applied_rules: HashMap<ExprId, HashSet<RuleId>>,
+    /// Transformation rules that may be used while exploring
+    /// (logical -> logical)
+    transformation_rules: Arc<[(RuleId, Arc<dyn Rule<T, Self>>)]>,
+    /// Implementation rules that may be used while optimizing
+    /// (logical -> physical)
+    implementation_rules: Arc<[(RuleId, Arc<dyn Rule<T, Self>>)]>,
     disabled_rules: HashSet<usize>,
     cost: Arc<dyn CostModel<T, M>>,
     property_builders: Arc<[Box<dyn PropertyBuilderAny<T>>]>,
@@ -94,29 +103,52 @@ impl Display for PredId {
 
 impl<T: NodeType> CascadesOptimizer<T, NaiveMemo<T>> {
     pub fn new(
-        rules: Vec<Arc<dyn Rule<T, Self>>>,
+        transformation_rules: Arc<[Arc<dyn Rule<T, Self>>]>,
+        implementation_rules: Arc<[Arc<dyn Rule<T, Self>>]>,
         cost: Box<dyn CostModel<T, NaiveMemo<T>>>,
         property_builders: Vec<Box<dyn PropertyBuilderAny<T>>>,
     ) -> Self {
-        Self::new_with_prop(rules, cost, property_builders, Default::default())
+        Self::new_with_prop(
+            transformation_rules,
+            implementation_rules,
+            cost,
+            property_builders,
+            Default::default(),
+        )
     }
 
     pub fn new_with_prop(
-        rules: Vec<Arc<dyn Rule<T, Self>>>,
+        transformation_rules: Arc<[Arc<dyn Rule<T, Self>>]>,
+        implementation_rules: Arc<[Arc<dyn Rule<T, Self>>]>,
         cost: Box<dyn CostModel<T, NaiveMemo<T>>>,
         property_builders: Vec<Box<dyn PropertyBuilderAny<T>>>,
         prop: OptimizerProperties,
     ) -> Self {
-        let tasks = VecDeque::new();
+        let tasks = Vec::new();
+        // Assign rule IDs
+        let transformation_rules: Arc<[(RuleId, Arc<dyn Rule<T, Self>>)]> = transformation_rules
+            .iter()
+            .enumerate()
+            .map(|(i, r)| (i, r.clone()))
+            .collect();
+        let implementation_rules: Arc<[(RuleId, Arc<dyn Rule<T, Self>>)]> = implementation_rules
+            .iter()
+            .enumerate()
+            .map(|(i, r)| (i + transformation_rules.len(), r.clone()))
+            .collect();
+        debug_assert!(transformation_rules.iter().all(|(_, r)| !r.is_impl_rule()));
+        debug_assert!(implementation_rules.iter().all(|(_, r)| r.is_impl_rule()));
         let property_builders: Arc<[_]> = property_builders.into();
         let memo = NaiveMemo::new(property_builders.clone());
         Self {
             memo,
+            task_counter: AtomicUsize::new(0),
             tasks,
             explored_group: HashSet::new(),
             explored_expr: HashSet::new(),
-            fired_rules: HashMap::new(),
-            rules: rules.into(),
+            applied_rules: HashMap::new(),
+            transformation_rules,
+            implementation_rules,
             cost: cost.into(),
             ctx: OptimizerContext::default(),
             property_builders,
@@ -128,7 +160,7 @@ impl<T: NodeType> CascadesOptimizer<T, NaiveMemo<T>> {
     /// Clear the memo table and all optimizer states.
     pub fn step_clear(&mut self) {
         self.memo = NaiveMemo::new(self.property_builders.clone());
-        self.fired_rules.clear();
+        self.applied_rules.clear();
         self.explored_group.clear();
         self.explored_expr.clear();
     }
@@ -153,8 +185,12 @@ impl<T: NodeType, M: Memo<T>> CascadesOptimizer<T, M> {
         self.cost.clone()
     }
 
-    pub fn rules(&self) -> Arc<[Arc<dyn Rule<T, Self>>]> {
-        self.rules.clone()
+    pub fn transformation_rules(&self) -> Arc<[(RuleId, Arc<dyn Rule<T, Self>>)]> {
+        self.transformation_rules.clone()
+    }
+
+    pub fn implementation_rules(&self) -> Arc<[(RuleId, Arc<dyn Rule<T, Self>>)]> {
+        self.implementation_rules.clone()
     }
 
     pub fn disable_rule(&mut self, rule_id: usize) {
@@ -215,7 +251,7 @@ impl<T: NodeType, M: Memo<T>> CascadesOptimizer<T, M> {
     /// Optimize a `RelNode`.
     pub fn step_optimize_rel(&mut self, root_rel: ArcPlanNode<T>) -> Result<GroupId> {
         let (group_id, _) = self.add_new_expr(root_rel);
-        self.fire_optimize_tasks(group_id)?;
+        self.fire_optimize_tasks(group_id);
         Ok(group_id)
     }
 
@@ -247,17 +283,30 @@ impl<T: NodeType, M: Memo<T>> CascadesOptimizer<T, M> {
         res
     }
 
-    fn fire_optimize_tasks(&mut self, group_id: GroupId) -> Result<()> {
-        trace!(event = "fire_optimize_tasks", root_group_id = %group_id);
-        self.tasks
-            .push_back(Box::new(OptimizeGroupTask::new(group_id)));
+    pub fn get_next_task_id(&self) -> usize {
+        self.task_counter.fetch_add(1, Ordering::AcqRel)
+    }
+
+    pub fn push_task(&mut self, task: Box<dyn Task<T, M>>) {
+        self.tasks.push(task);
+    }
+
+    fn pop_task(&mut self) -> Option<Box<dyn Task<T, M>>> {
+        self.tasks.pop()
+    }
+
+    fn fire_optimize_tasks(&mut self, root_group_id: GroupId) {
+        trace!(event = "fire_optimize_tasks", root_group_id = %root_group_id);
+        let initial_task_id = self.get_next_task_id();
+        self.push_task(get_initial_task(initial_task_id, root_group_id));
         // get the task from the stack
         self.ctx.budget_used = false;
         let plan_space_begin = self.memo.estimated_plan_space();
         let mut iter = 0;
-        while let Some(task) = self.tasks.pop_back() {
-            let new_tasks = task.execute(self)?;
-            self.tasks.extend(new_tasks);
+        while let Some(task) = self.pop_task() {
+            task.execute(self);
+
+            // TODO: Iter is wrong
             iter += 1;
             if !self.ctx.budget_used {
                 let plan_space = self.memo.estimated_plan_space();
@@ -286,12 +335,11 @@ impl<T: NodeType, M: Memo<T>> CascadesOptimizer<T, M> {
                 }
             }
         }
-        Ok(())
     }
 
     fn optimize_inner(&mut self, root_rel: ArcPlanNode<T>) -> Result<ArcPlanNode<T>> {
         let (group_id, _) = self.add_new_expr(root_rel);
-        self.fire_optimize_tasks(group_id)?;
+        self.fire_optimize_tasks(group_id);
         self.memo.get_best_group_binding(group_id, |_, _, _| {})
     }
 
@@ -374,15 +422,15 @@ impl<T: NodeType, M: Memo<T>> CascadesOptimizer<T, M> {
         self.explored_expr.remove(&expr_id);
     }
 
-    pub(super) fn is_rule_fired(&self, group_expr_id: ExprId, rule_id: RuleId) -> bool {
-        self.fired_rules
+    pub(super) fn is_rule_applied(&self, group_expr_id: ExprId, rule_id: RuleId) -> bool {
+        self.applied_rules
             .get(&group_expr_id)
             .map(|rules| rules.contains(&rule_id))
             .unwrap_or(false)
     }
 
-    pub(super) fn mark_rule_fired(&mut self, group_expr_id: ExprId, rule_id: RuleId) {
-        self.fired_rules
+    pub(super) fn mark_rule_applied(&mut self, group_expr_id: ExprId, rule_id: RuleId) {
+        self.applied_rules
             .entry(group_expr_id)
             .or_default()
             .insert(rule_id);
@@ -406,3 +454,18 @@ impl<T: NodeType, M: Memo<T>> Optimizer<T> for CascadesOptimizer<T, M> {
         self.get_property_by_group::<P>(self.resolve_group_id(root_rel), idx)
     }
 }
+
+pub fn rule_matches_expr<T: NodeType, M: Memo<T>>(
+    rule: &Arc<dyn Rule<T, CascadesOptimizer<T, M>>>,
+    expr: &ArcMemoPlanNode<T>,
+) -> bool {
+    let matcher = rule.matcher();
+    let typ_to_match = &expr.typ;
+    match matcher {
+        RuleMatcher::MatchNode { typ, .. } => typ == typ_to_match,
+        RuleMatcher::MatchDiscriminant {
+            typ_discriminant, ..
+        } => *typ_discriminant == std::mem::discriminant(typ_to_match),
+        _ => panic!("IR should have root node of match"), // TODO: what does this mean? replace text
+    }
+}

optd-core/src/cascades/tasks.rs

@@ -3,26 +3,31 @@
 // Use of this source code is governed by an MIT-style license that can be found in the LICENSE file or at
 // https://opensource.org/licenses/MIT.
 
-use anyhow::Result;
 
-use super::{CascadesOptimizer, Memo};
+use super::{CascadesOptimizer, GroupId, Memo};
 use crate::nodes::NodeType;
 
 mod apply_rule;
+mod explore_expr;
 mod explore_group;
-mod optimize_expression;
+mod optimize_expr;
 mod optimize_group;
 mod optimize_inputs;
 
-pub use apply_rule::ApplyRuleTask;
-pub use explore_group::ExploreGroupTask;
-pub use optimize_expression::OptimizeExpressionTask;
 pub use optimize_group::OptimizeGroupTask;
-pub use optimize_inputs::OptimizeInputsTask;
 
 pub trait Task<T: NodeType, M: Memo<T>>: 'static + Send + Sync {
-    fn execute(&self, optimizer: &mut CascadesOptimizer<T, M>) -> Result<Vec<Box<dyn Task<T, M>>>>;
+    fn execute(&self, optimizer: &mut CascadesOptimizer<T, M>);
+}
 
-    #[allow(dead_code)]
-    fn describe(&self) -> String;
+pub fn get_initial_task<T: NodeType, M: Memo<T>>(
+    initial_task_id: usize,
+    root_group_id: GroupId,
+) -> Box<dyn Task<T, M>> {
+    Box::new(OptimizeGroupTask::new(
+        None,
+        initial_task_id,
+        root_group_id,
+        None,
+    ))
 }