Documentation / Architecture

High-performance code intelligence system in Rust. Indexes code, tracks relationships, serves via MCP.

1. Parse fast - Tree-sitter AST parsing (same as GitHub code navigator) for Rust, Python, TypeScript, JavaScript, Java, Kotlin, Go, PHP, C, C++, C#, Swift, and GDScript
2. Extract real stuff - functions, traits, type relationships, call graphs
3. Embed - semantic vectors built from your doc comments
4. Index - Tantivy + memory-mapped symbol cache for <10ms lookups
5. Serve - MCP protocol for AI assistants, ~300ms response time (HTTP/HTTPS) and stdio built-in (0.16s)

• How It Works - Detailed system architecture
• Memory Mapping - Cache and storage design
• Embedding Model - Semantic search implementation
• Language Support - Parser system and adding languages

Parallel indexing pipeline: 5-stage architecture (DISCOVER → READ → PARSE → COLLECT → INDEX) with work-stealing queues. Phase 2 runs EmbeddingPool for parallel embedding generation.

Memory-mapped storage: Vector cache for semantic search:

• segment_0.vec - 384-dimensional vectors, <1μs access after OS page cache warm-up

Embedding lifecycle management: Old embeddings deleted when files are re-indexed to prevent accumulation.

Lock-free concurrency: DashMap for concurrent reads, RwLock for Tantivy writes.

IndexFacade: Unified interface wrapping DocumentIndex, Pipeline, and SemanticSearch.

Language-aware semantic search: Embeddings track source language, enabling filtering before similarity computation. No score redistribution - identical docs produce identical scores regardless of filtering.

Hot reload: File watcher with 500ms debounce triggers re-indexing of changed files only.

Parser benchmarks on a 750-symbol test file:

Run performance benchmarks:

codanna benchmark all          # Test all parsers
codanna benchmark python       # Test specific language

• Learn about User Guide for usage
• Explore Advanced features
• Read Contributing to add features

Back to Documentation