Phase 4: Query & Quality (Months 8-9)

Goals: Unified query engine, testing improvements

Duration: 2 months
Dependencies: Phase 3 Complete

Objectives

Month 8: Query Engine Architecture

• Design unified query pipeline (LogicalPlan/PhysicalPlan abstraction)
• Implement logical plan classes (Scan, Filter, Project, Join, Aggregate, Sort, Limit)
• Implement physical operators (ScanOperator, FilterOperator, ProjectOperator, JoinOperators, etc.)
• Implement query optimizer (rule-based and cost-based)
• Refactor SQL engine to use new pipeline
• Add query plan caching
• Optimize top 10 query patterns (20% performance improvement target)

Month 9: Testing Infrastructure

• Create test base classes (ArcadeDBTestBase, GraphTestBase, QueryEngineTestBase, etc.)
• Implement test data builders (fluent API for test data creation)
• Add mutation testing (PIT integration)
• Increase test coverage to 60%
• Create performance benchmarking suite (50+ JMH benchmarks)
• Add property-based testing (jqwik)

Query Pipeline Design

Parse → LogicalPlan → Optimize → PhysicalPlan → Execute → ResultSet

Logical Plan Nodes: Scan, Filter, Project, Join, Aggregate, Sort, Limit
Physical Operators: IndexScan, TableScan, FilterOperator, ProjectOperator, NestedLoopJoin, HashJoin, HashAggregate, SortOperator, LimitOperator

Deliverables

• Unified query execution framework
• All query languages using same pipeline
• Query optimizer with 10+ optimization rules
• Comprehensive test infrastructure
• 60%+ code coverage
• Performance benchmarking suite (50+ benchmarks)

Success Metrics

• Query performance improved by 20%
• Test coverage: 60%+
• Mutation coverage: 70%+
• 50+ JMH benchmarks running
• Plan caching with 50%+ hit rate

Related Tasks

Tasks TASK-P4-001 through TASK-P4-013 will be created as individual issues linked to this epic.

Part of the ArcadeDB Evolution Project - see IMPROVEMENT_PLAN.md