oxc-project__oxc/crates/oxc_minifier/docs/ARCHITECTURE.md

Architecture

Design Philosophy

Achieve maximum compression through comprehensive optimizations while maintaining correctness.

Size Optimization Strategy

• Fixed-point iteration until no more improvements
• Combine multiple optimization techniques
• Learn from all major minifiers
• Apply optimizations exhaustively

Learning from Industry Leaders

From Closure Compiler (Size Focus)

• Advanced dead code elimination
• Aggressive constant folding
• Cross-statement optimizations
• Property flattening techniques
• Whole program analysis

From Terser/UglifyJS (Comprehensive)

• All peephole optimizations
• Battle-tested patterns
• Edge case handling
• Extensive transformation catalog

From esbuild (Performance)

• Minimal AST passes where possible
• Arena allocation strategy
• Efficient algorithms
• Smart traversal patterns

From SWC (Modern)

• Rust safety and performance
• Clean visitor pattern
• Parallel processing potential
• Modern architecture

Core Components

Compressor (compressor.rs)

Orchestrates the optimization pipeline with fixed-point iteration.

pub struct Compressor<'a> {
    allocator: &'a Allocator,
}

Key responsibilities:

• Builds semantic model
• Applies normalization
• Runs peephole optimizations to fixed-point
• Manages optimization state

Peephole Optimizations (peephole/)

17+ transformation passes including:

• Constant folding
• Dead code elimination
• Control flow optimization
• Expression simplification
• Syntax substitution

Each optimization implements transformations in the AST traversal visitor pattern.

Context (ctx.rs)

Shared utilities for optimizations:

• AST manipulation helpers
• Constant evaluation
• Side effect analysis
• Semantic utilities

pub struct Ctx<'a, 'b> {
    // Provides access to:
    // - AST builder
    // - Scoping information
    // - Symbol values
    // - Optimization options
}

State Management (state.rs)

Tracks optimization state:

• Symbol values
• Changed flags
• Pure functions
• Source type

Optimization Pipeline

1. Parse and Build Semantic Model
   └─> SemanticBuilder creates scoping and symbols

2. Normalization Pass
   └─> Convert to canonical form for optimizations

3. Peephole Optimization Loop
   └─> Apply all optimizations
   └─> Check if any changes made
   └─> Repeat until fixed-point (no changes)

4. Optional Mangling
   └─> Rename variables for size

5. Code Generation
   └─> Output optimized JavaScript

Memory Management

• Arena allocation via oxc_allocator
• Minimal cloning through TakeIn trait
• Efficient AST manipulation with arena pointers

Traversal Strategy

Using oxc_traverse for AST walking:

• Enter/exit hooks for each node type
• Bottom-up transformations (exit handlers)
• State tracking through context

Integration Points

• oxc_ast: AST node definitions
• oxc_semantic: Scope and symbol analysis
• oxc_ecmascript: ECMAScript operations
• oxc_mangler: Variable renaming
• oxc_codegen: Output generation