- [ ] Add notes on the architecture of the reference implementation - [ ] Explain - [ ] Parsing entire input - [ ] Threading the tokens - [ ] Reserved word discovery - [ ] Dealing with gluing and line-breaks as soft semi-colons - [ ] Hand-crafted tokeniser to deal with: - [ ] Multi-line strings - [ ] Interpolated strings - [ ] Gluing - [ ] Line breaks - [ ] Discovery of reserved words - [ ] Hand-crafted parser based on recursive descent with operator precedence - [ ] Key idea: form context vs expression context