Static Semantics Definition with Statix

Programs that are syntactically well-formed are not necessarily valid programs. Programming languages typically impose additional context-sensitive requirements on programs that cannot be captured in a syntax definition. Languages use names to identify reusable units that can be invoked at multiple parts in a program. In addition, statically typed languages require that expressions are consistently typed. The Statix language supports the specification of name binding and type checking rules of a language. The rules of the static semantics are written as logic rules, and solved using a constraint-based approach, and uses scope graphs for name resolution.

Table of Contents

• 1. Getting Started
• 2. Debugging
  • 2.1. Basic Checklist
  • 2.2. Checking AST Traversal
  • 2.3. Checking Reference Resolution
  • 2.4. Interpreting Error Messages
  • 2.5. Inspecting Variables
  • 2.6. Inspecting the Scope Graph
  • 2.7. Creating Minimal Examples
  • 2.8. Testing Predicates
  • 2.9. Some Common Problems
  • 2.10. Getting Help and Reporting Issues
• 3. Language Reference
  • 3.1. Lexical matters
  • 3.2. Terms
  • 3.3. Modules and Tests
  • 3.4. Signatures
  • 3.5. Predicates and Rules
  • 3.6. Constraints
• 4. Stratego API
• 5. Signature Generator
  • 5.1. Well-Formed SDF3 Requirements
  • 5.2. Applying the Generator in Spoofax 2
  • 5.3. Using the Generated Injection strategies
  • 5.4. Using the Generated Signatures
  • 5.5. Troubleshooting
• 6. Rename Refactoring
  • 6.1. Renaming in NaBL2
• 7. NaBL2 Migration Guide
  • 7.1. Terms
  • 7.2. Name Resolution
  • 7.3. Functions
  • 7.4. Relations
• 8. Migrating to the Concurrent Solver
  • 8.1. Enabling the Concurrent Solver
  • 8.2. Indirect Type Declaration
  • 8.3. Using new Solver Features