Compiler Pipeline

Source → Lexer → Parser → Semantic Analyzer → IR Lowering → (Plugin System)
           │        │            │                 │               │
           ▼        ▼            ▼                 ▼               ▼
        Tokens     AST     Validated AST       IrModule      IrPass / Backend

• Lexer: Tokenizes source with logos.
• Parser: Builds AST from tokens with chumsky (Pratt precedence).
• Semantic Analyzer: 6-pass validation (Pass 0 resolves modules, Passes 1–5 build symbol tables, resolve types, validate expressions, validate traits, detect cycles). Inference, validation, and the per-function scratch state (local_let_bindings, inference_scope_stack) operate on SemType, a structural type representation. There are no string sentinels ("Unknown", "InferredEnum", "Nil") inside the analyzer; callers gate on SemType::Unknown / is_indeterminate() directly. The SymbolTable::LetInfo::inferred_type slot is the only string-typed storage that remains — it is the public boundary with IR lowering and external consumers (LSP queries, downstream tooling).
• IR Lowering: Converts the validated AST + symbol table into a fully type-resolved IrModule. Module nesting is flattened in the per-type vectors: inline mod foo { struct Bar { ... } } lowers to a top-level IrStruct { name: "foo::Bar", ... }, so backends that don't care about source structure see a flat list of definitions keyed by qualified name. A parallel IrModule.modules: Vec<IrModuleNode> tree mirrors the source mod hierarchy with per-module ID lists for backends that need namespaced output.
• Plugin System: External IrPass transforms and Backend emitters composed through Pipeline: see Plugin System.

Compiler Outputs

See the AST Reference and IR Reference for the data shapes each phase produces.