pierrelf.com
Implement a JavaScript parser that transforms a token stream into an Abstract Syntax Tree (AST).
Scope#
Build a recursive-descent parser for ECMAScript 2024. The parser should handle the full expression grammar and all major statement/declaration types.
AST Node Types#
Expressions:
- Literals: number, string, boolean, null, undefined, array, object, regex
- Identifier, ThisExpression
- Unary, binary, logical, conditional (ternary) expressions
- Assignment expressions (=, +=, -=, etc.)
- Member access (dot and bracket)
- Call expressions and new expressions
- Arrow function expressions
- Spread element
- Template literal expressions
- Comma expressions
Statements:
- ExpressionStatement, BlockStatement
- VariableDeclaration (var/let/const)
- IfStatement, SwitchStatement
- ForStatement, ForInStatement, ForOfStatement
- WhileStatement, DoWhileStatement
- ReturnStatement, ThrowStatement
- TryStatement (try/catch/finally)
- BreakStatement, ContinueStatement
- LabeledStatement
Declarations:
- FunctionDeclaration (including generator
function*) - ClassDeclaration (constructor, methods, static methods, getters/setters)
- Import/Export declarations (basic forms)
Features#
- Operator precedence handling (Pratt parsing or precedence climbing)
- Destructuring patterns in variable declarations and parameters
- Default parameter values
- Rest parameters (
...args) - Computed property names (
[expr]: value) - Shorthand property/method definitions in object literals
- Syntax error reporting with source location
Acceptance Criteria#
- AST type definitions for all node types
- Parser produces correct AST for expressions with proper precedence
- All statement types parsed correctly
- Function and class declarations parsed
- Destructuring patterns supported
- Arrow functions parsed (including concise body)
- Meaningful syntax error messages with line/column
- Unit tests for each grammar production
Phase 10 — JavaScript Engine (issue 2 of 15). Depends on: JS Lexer/Tokenizer.